Сommodity and Money Flows in Russian Empire in the Second half of the Nineteenth Century: Network and Geoinformation Analysis

Salomatina Sof'ya ORCID: 0000-0003-0748-6229 PhD in History Associate Professor, Department of Historical Information Science, Faculty of History, Lomonosov Moscow State University 119192, Russia, Moscow, Lomonovsky Prospekt, 27, building 4, of. G423 ssalomatina@gmail.com Other publications by this author

IntroductionOn the movement of commodity-money flows in the Russian Empire in the second half of the XIX century.

it is possible to understand the distribution of resources in the economy from the point of view of geography and hierarchy of internal markets. This approach makes it possible to identify the uneven development of regions, which could affect the well-being of the population. However, to study the domestic markets of the XIX century, there are often not enough nationwide comparable sources for a sufficiently long period. This gap is filled by the statistics of transfers between branches of the State Bank of the Russian Empire (hereinafter the State Bank), which was published in its annual reports until the beginning of the XX century. These statistics reflect payments for goods and services, and this cash flow is counter-directed to the commodity. Since in the second half of the XIX century, the State Bank was the largest commercial bank with a network of branches throughout the country, its cash flows can be used in the most general way to identify the main commodity flows in the country.

This article is a continuation of the article by S. A. Salomatina, I. M. Garskova and T. Ya. Valetov on settlements related to the 10 largest financial centers of the Russian Empire in 1898 (St. Petersburg, Moscow, Odessa, Kiev, Warsaw, Riga, Baku, Nizhny Novgorod, Kharkov, Rostov-on-Don) ^[10]. In that article it was shown that the payment network of the State Bank consisted of two segments. The highest, or capital, included payment flows in which one side was St. Petersburg or Moscow, the rest of the flows can be attributed to the regional segment.

Taking into account these developments, the new article examines two issues. Firstly, what changes have occurred in the top segment of the payment network, around St. Petersburg and Moscow, from the 1860s to the 1890s? Secondly, how the geography of the largest regional cash flows has changed over these 30 years. To answer these questions, a new data set was created, which included all the statistics of State Bank transfers for 1868, 1878, 1888 and 1898. If 1898 is the last full year of the economic recovery of the 1890s, before the crisis of 1899-1903, then three more time slices allow us to assess the situation in previous decades in increments at the age of 10.

Cash flow statistics are processed by network analysis methods, in which each flow between two branches is considered as a unidirectional connection. The connections were divided into metropolitan and regional ones, which made it possible to separately trace the dynamics of the two segments of the payment network. Network graphs are represented on geographical maps, and the nodes of the graphs correspond to the settlements in which the branches of the State Bank were located. The visualization of the network is made separately for the European and Asian parts of the empire, taking into account the railways and main waterways.

The statistics of transfers of the State Bank are used in another of our articles on outgoing payment flows (sales of goods) from the agrarian Oryol province, i.e. from the Oryol and Yelets branches of the State Bank ^[12]. In that study it was shown that the priority direction of export from the Orel province since the 1870s was western, to the Baltic ports, along the Riga— Tsaritsyn highway. Then, as the railway network expanded, this cash flow shifted south and west. The priority of the western direction of the export of agricultural products from the Oryol province is confirmed by data on payments serviced by the Oryol Commercial Bank ^[11]. Thus, this regional example can be considered as a detail to some of the plots of this new article.

As already noted, similar methods are used in economic geography to rank cities by various parameters, as well as to visualize their spatial relationships ^{[10] [19] [15] [17] [18]}. In addition, there is already a fairly traditional topic in financial history related to the study of financial centers, which are arranged hierarchically from regional markets to national and further to international ^{[16] [13] [4] [8]}.

The article has the following structure: the first section contains a comment on the source used. The second section analyzes the generalized statistics of transfers between branches of the State Bank for 1868, 1878, 1888 and 1898. The third section is devoted to the evolution of the higher (metropolitan) segment of the payment network. The fourth section presents the analysis and visualization of network graphs on geographical maps for regional commodity and cash flows.

Thus, this article analyzes the question of the structure of the national payment network, the ratio of the metropolitan and regional segments in it, changes in this system over more than 30 years — all this is the key to understanding the mechanism of redistribution of funds in the economy of the Russian Empire as a whole. Sources and database used

The statistics of transfers between branches of the State Bank have already been described in detail in our articles mentioned above, so now it is enough to note the most important features of these data.

The annual reports of the State Bank from 1861 to 1900 published annual amounts of transfers from each branch to other branches, which can be considered as unidirectional connections. Each such relationship is an accounting unit (record) in the database. For example, 1.3 million rubles were sent from Moscow to Irkutsk in 1898. Transfers in the opposite direction, from Irkutsk to Moscow, amounted to 2.4 million rubles. — this is another connection and, accordingly, an entry in the database.

Our study uses all the data on transfers for 1868, 1878, 1888, 1898. ^{[1] [2] [3] [9, c. 57–81]}. The number of records in such a database, of course, depends on the number of State Bank branches, which were 47 in 1868, 59 in 1878, 97 in 1888, 122 in 1898 (Table 1b). So, in 1868 there should be 47 ? 47 = 2,209 entries, including intra-city transfers, when the place of sending and receiving the transfer coincide. However, in fact, there were only 929 connections in the source for 1868, because there were no transfers between some branches that year, for example, between Kazan and Vologda or between Ryazan and Astrakhan, which in itself indicates that economic ties are unevenly distributed across the country. As a result, there are 17,116 records in the database for four time slices, and not 29,983 if each branch had bidirectional connections with all other branches and a flow of intra-city transfers.

The statistics of State Bank transfers have features that are not noted in our previous articles. Firstly, there is insufficient data on transfers of the early period related to Poland; these data are comparable only since 1886. The fact is that until 1885 a Polish bank operated in this region with a board in Warsaw and a network of branches (in 1868 in Lodz and W?oc?awek, since 1876 10 branches — the two above mentioned, as well as Kalisz, Kielce, Lomzha, Lublin, Petrokov, Plock, Radom, Czestochowa). Statistics of transfers by a Polish bank, similar to the State Bank, have not yet been found, there is a high probability that it does not exist. Only after the liquidation of the Polish Bank in 1885, the Warsaw office of the State Bank and 10 branches in the same cities were opened. Since 1886, full-fledged translation statistics have been published on these points. For an earlier period, since 1875, the reports of the State Bank have data on transfers to Warsaw from the branches of the State Bank. However, the first branch of the State Bank in Poland opened in 1884 in Tomashev, apparently this was due to the royal hunting residence in the nearby Spala, which began to expand just in those years. The Tomashev branch is reflected in the statistics of transfers since 1884 ^{[14, p. 199]}

Secondly, the statistics of the State Bank have data on branches at the largest fairs, but there are limited opportunities to study the Nizhny Novgorod Fair. As you know, some branches of the State Bank had a temporary status, i.e. they did not operate all year. Such branches could be opened at fairs during their work or at resorts during the season (for example, Irbit in Perm province or Yalta in Taurida province). Temporary branches are taken into account in our calculations as well as permanent ones, because there is no difference between them when working with transfers. However, in 1868 a permanent branch was opened in Nizhny Novgorod, while a fair office existed here since 1820 from the State Commercial Bank, the direct predecessor of the State Bank ^{[7, p. 6-14]}. The report for 1868 published data on two branches at once (permanent and temporary), which are summarized in our calculations (for sending and receiving). However, since 1869, in the reports of the State Bank, operations in Nizhny Novgorod are shown as if there was only one permanent branch operating there, which also served the fair, while there is no separate data on the fair. A similar situation in 1868 was in Poltava, where a permanent branch was opened with a previously existing temporary one.

Thirdly, intra-city transfers are a new phenomenon in the late 1890s. Cash flows that have the same place of sending and receiving appear only since 1897, apparently, they were simply not taken into account before, although it has not yet been possible to find primary data on transfer statistics to confirm this hypothesis. Intra-city transfers could be significant. For example, in 1898, transfers within Moscow amounted to 1.7 million rubles, and within Nizhny Novgorod — 1.6 million rubles. Such figures are excluded from calculations because they do not reflect interregional ties.

Thus, this study analyzes the statistics of directed cash flows between geographical points, the list of which is clearly defined and is gradually expanding from 1868 to 1898. Further, the overall growth of these payment networks is considered. Transfers between branches of the State Bank: generalized indicators

The growth of the network of interregional transfers of the State Bank from 1868 to 1898 is presented in tables 1a and 1b in millions of rubles and by the number of unidirectional connections between branches.

The annual volume of transfers during this period increased 4.7 times — from 287.0 to 1,350.9 million rubles. On the one hand, this is due to the growing number of branches. On the other hand, new branches appear in response to the growing demand for banking services. If we assume that economic reasons were not the main ones when opening a branch, this would have to be reflected in the weak cash flows generated by this branch.

Let's take a closer look at how this increase in calculations was distributed between capitals and regions. In metropolitan communications, at least one party (by sending or receiving) is St. Petersburg or Moscow. In absolute terms, the capital's communications grew rapidly in the early period, i.e. by 1878 — from 237.0 to 550.1 million rubles. There were fewer new branches then than in the following decades — only 12 (Table 1b), so most likely this growth is largely due to the development of communications, primarily railways. Later, until the end of the XIX century. the metropolitan segment of the payment network grew more slowly, but by 1898 it amounted to 760.4 million rubles. In relative terms, metropolitan communications accounted for 82.6% of all transfers in 1868, and in 1898 — less, 56.3%, but still more than half of all flows. Thus, it is also necessary to take into account the growth of the payment network outside the metropolitan segment. Table 1a.

Transfers between branches of the State Bank, in millions of rubles.In millions of rubles.

 Table 1b. Transfers between State Bank branches, number of connections1868

* — Without intra-city transfers, in which the place of departure and receipt coincided (in the same branch).
** — At least one side of the capital's relations is St. Petersburg or Moscow.
*** — There are no St. Petersburg and Moscow in regional relations.Regional connections, i.e. transfers between all branches except St. Petersburg and Moscow, in 1868 amounted to only 50 million rubles, but by 1898 they had increased 11.8 times to 590.5 million rubles.

In relative terms, regional transfers increased from 17.4% of the entire payment network in 1868 to 43.7% in 1898, i.e. the regional segment was gradually catching up with the metropolitan one.

The growth processes of payment networks can also be represented through the number of connections (Table 1b). Recall that one connection is transfers between two branches in the same direction. The number of connections depends on the number of branches, this is what limits the number of metropolitan connections (St. Petersburg and Moscow), because the capitals, as a rule, maintained contacts with almost all existing branches, whereas for regional offices this was far from the case. As a result, the overall growth in the number of connections was largely determined by the development of the regional segment — from 745 in 1868 to 9,284 in 1898. The growth rate in this case was higher than the growth rate of the number of branches, because new cash flows appeared between those pairs of branches that previously had no connections.

Thus, even in general figures, the formation of a two-tier payment network is noticeable: the metropolitan and regional segments. This division is not accidental. The capital segment receives cash flows associated with the main national capital markets, which play a crucial role in the integration of the country's regional financial markets. The following section is devoted to the capital segment.

Evolution of the highest (metropolitan) segment of the payment networkCapital financial markets link regional markets at a higher level of the hierarchy.

Monetary resources, as well as the supply of goods and services, are concentrated in the capital's markets, where regions can meet their demand. Because of this concentration of connections, metropolitan markets are much larger than regional ones. Due to the disparate sizes of flows, the metropolitan and regional segments need to be analyzed separately, otherwise regional flows are "lost" against the background of the capital. In our article about the leading Russian financial centers, there is a visualization of a part of the capital segment of 1898, where it is shown that these are two networks of the "star" type, in which connections (cash flows) exit and originate from the center (St. Petersburg or Moscow) ^[10]. Graphs for other time slices in the most general form looks similar. However, a more detailed analysis of the metropolitan segment from 1868 showed that in fact it is a "double star", as described below.

When analyzing the "stars", it is necessary to divide the streams into outgoing (the core as the place of sending transfers) and incoming (the core as the place of receiving transfers). Sending transfers from a financial center corresponds to the purchase of goods and services by this center for consumption or redistribution (resale). Receiving transfers corresponds to the sale of goods and services from this center to other regions. Unfortunately, it is impossible to detail the structure of these flows from archival sources, because the archives of the State Bank in St. Petersburg (the Russian State Historical Archive) and Moscow (the Central State Archive of Moscow) store data of the same level of aggregation as those published in the reports of the State Bank used in this study. At the same time, no evidence has yet been found that these flows include transfers at the expense of the treasury, i.e. it is assumed that these are only customer transfers (commercial transfers).

The statistics of metropolitan payment networks on payments for purchases and sales are presented in Tables 2a and 2b. First of all, it is necessary to pay attention to the flows between the two capitals, which, due to the huge volume, are allocated in a separate column. These streams make one double star out of two stars. This double star looks about the same for 1868, 1878 and 1888, despite the difference in details, so as an example, Figure 1 shows a diagram for 1888. Then by 1898, the double star changes slightly, which is reflected in Figure 2. Moscow intensively buys in St. Petersburg: in 1868 by 47 million rubles, in 1878 by 74.6 million rubles, in 1888 by 39.9 million rubles. The return flow (St. Petersburg buys in Moscow) is small and quite comparable with the metropolitan connections of other major cities in Russia (2.4, 8.7 and 8.9 million rubles, respectively). By the end of the XIX century, the flows connecting the two capitals leveled off (15.1 million rubles from Moscow and 22.7 million rubles from St. Petersburg). If St. Petersburg used to buy little in Moscow, then in 1898 the situation changed. Another feature of the double star is the huge purchases of St. Petersburg throughout the country, which have always exceeded sales several times. Moscow's incoming and outgoing flows are generally smaller and more balanced. Purchases in Moscow were in the lead in 1868 and 1888, and in 1878 and 1898, on the contrary, sales.

Table 2a. Saint Petersburg and Moscow as a place of sending transfers (purchase of goods and services), mln rubles.Year

 Table 2b. St. Petersburg and Moscow as a place of receiving transfers (sale of goods and services), million rubles.Year

* — For flows of more than 100 thousand rubles. no traffic flows between St. Petersburg and Moscow.
** — Without intra-city transfers.
*** — No traffic flows between St. Petersburg and Moscow.As a result, if purchases greatly exceeded sales in St. Petersburg, this may indicate that a significant part of goods and services arrived in the capital, where they were consumed or exported, as well as a significant flow left St. Petersburg to Moscow, where it was redistributed across the country.

Regarding Moscow, where sales prevailed in 1878 and 1898, and in general incoming outgoing flows were close in volume, it is worth paying attention to the results of our article on the leading financial centers of the Russian Empire, which shows that in 1898 this was the situation in all major regional financial centers (Odessa, Kiev, Warsaw, Riga, Baku, Nizhny Novgorod, Kharkiv, Rostov-on-Don) ^[10]. The difference between Moscow and other centers is purely quantitative. If the annual sales volume in Odessa, Kiev, Warsaw was 50-60 million rubles a year, and in the other mentioned centers even less — 25-40 million rubles, then in Moscow this figure reached 200 million rubles. In this sense, Moscow was a higher-level sales center.

Figure 1. Capital segment of the payment network of the State Bank in 1888

Figure 2. Capital segment of the payment network of the State Bank in 1898 Evolution of regional payment networks from 1868 to 1898

Outside the metropolitan segment, at a lower level of the hierarchy, regional markets functioned, which are important to study in order to clarify the internal trade flows in the country.

At the same time, the "near" connections of the regional center usually concerned neighboring provinces, and the "far" connections depended on the location of communications, for example, Nizhny Novgorod and Baku are connected along the Volga basin, while Riga and Orel are located on the railway line from the Central Chernozem region to the Baltic ports.

Visualization of regional flows is presented in Figures 3-7 for 1868, 1878, 1888 and 1898. The network graphs are located on geographical maps, which represent all the settlements in which there were branches of the State Bank, although not all of them were signed due to lack of space. In addition, railways and navigable sections of the main rivers are marked on the maps so that it is possible to correlate the routes of transportation of goods with bank settlements. To create the maps, the "Map of steamship communications, railways and postal roads of the Russian Empire" (St. Petersburg, Edition of the Cartographic institution of A. Ilyin, 1911), provided by the Department of Cartographic Publications of the Russian State Library, was used. This map shows separately the European part of the empire and the south of the Asian part.

As already mentioned, regional flows can be analyzed if we exclude the highest or metropolitan level of the payment network, i.e. connections in which at least one side is St. Petersburg or Moscow. The graphs show only large regional flows. First of all, flows of more than 1 and 3 million rubles were allocated, this level of cutting off small flows was established in our article on financial centers in 1898. ^[10] This approach made visualization convenient, allowed comparing regional markets, and in most cases large flows made up a significant part of all calculations of a particular financial center. The cut-off level of 1 million rubles was preserved for comparison in this study, but it turned out to be too high for 1868, as well as for the Asian part of the Russian Empire, where flows of 0.5–1 million rubles are also represented.

Figure 3 shows the payment flows of 1868, when there were still very few operating railways, so to a large extent these are payments for goods transported by water and horse-drawn transport. In addition, it is worth noting once again that there is no data for Poland for 1868. 745 regional flows were reduced to 692 after the unification of flows associated with permanent and temporary branches in Nizhny Novgorod and Poltava. Only 18 flows exceeded 0.5 million rubles per year, but in total they accounted for 54%, i.e. more than half, of all non-capital flows.

In the south, Odessa's network stood out as a center of international maritime trade. It was here that the strongest regional ties were located in 1868 — payments from Odessa to Rostov-on-Don (5.7 million rubles) and to Taganrog (4.8 million rubles), which implied the accumulation of goods in these centers and then their export to Odessa. In this case, we are talking not only about the transportation of goods from the southern regions to the Black Sea ports. Rostov-on-Don had traditional ties with the center of European Russia, on the graph — with Nizhny Novgorod, through the Volgodonsk trans-shipment (Tsaritsynsky ferry) between the Volga and the Don, i.e. through the land crossing from the Volga to the Black Sea basin. The Volga-Don Railway has been operating on this section since 1862. Another very large connection of the Odessa hub is the flow from Tiflis (5.2 million rubles), which corresponds to the delivery of goods from Odessa to Transcaucasia via the Black Sea. Figure 3. Transfers between branches of the State Bank in 1868, with the exception of St. Petersburg and Moscow, cash flows of more than 0.5 million rubles. 

In the center of European Russia, a small knot around Nizhny Novgorod is noticeable.

These flows can be associated with a fair where goods were purchased for Siberia and for the south of the European part of the empire. The cash flow from Yaroslavl to Samara indicates the flow of goods in the opposite direction, up the Volga, and then through the canal systems to St. Petersburg. Agricultural products were transported along this route.

The export of goods along the northern rivers to Arkhangelsk was primarily from Vyatka province, despite the fact that access to the White Sea basin in this area was possible only by horse-drawn transport from the north of Vyatka to the south of Vologda province ^{[6, p. 165-176]}. Interestingly, the flow of goods along the rivers from Vologda to Arkhangelsk was smaller in volume than from Vyatka.

In Siberia, the largest commodity exchange was observed between Tomsk and Irkutsk, and the export of goods from Irkutsk (cash flow to Irkutsk) dominated here. Transportation was carried out along the main highway — the Moscow-Siberian Highway ^[5], along which it was possible to get to both the Nizhny Novgorod and Irbit fairs.

Thus, in 1868 there were almost no regional connections due to the development of railway transport, so it is important to assess the changes 10 years later, in 1878, when the main railway network in the European part of the empire was formed. In Figure 4, the payment flows of 1878 are divided into two fragments — for the European and Asian parts of the country. In total, 43 flows out of 1,490, or 45% in monetary terms of all regional flows for 1878, were visualized.

For European Russia, only flows of more than 1 million rubles are shown (the left part of Figure 4). These are the 24 largest streams, in the geography of which there is a noticeable difference since 1868. The new "railway" connections are significant commodity—money flows along the Riga-Tsaritsyn highway (Smolensk, Orel, Borisoglebsk), and the Riga—Orel flow was 9.0 million rubles (the third largest regional flow in 1878). All these flows are associated with the export of agricultural products from the Central Chernozem region to the Baltic ports. Kursk also gravitated to this highway, although the flow of goods to Kiev from there was also significant.

By 1878, the Odessa junction was strengthened, and precisely due to connections with southern cities, where many new railways appeared — with Kiev, Kharkov, Poltava, Yekaterinoslav. At the same time, cash flows dominated in Odessa's connections, corresponding to the import of goods from Nikolaev (13.5 million rubles, the largest flow in 1878), Taganrog (6.1 million rubles), Rostov-on-Don (4.9 million rubles), Kiev (2.7 million rubles). A similar process took place in the Kiev node — communications with Odessa, Warsaw, Kharkov, Kursk.

Warsaw is represented in Figure 4 only by transfers from the branches of the State Bank to the main office of the Polish Bank in Warsaw, i.e. in this case only half of the data is available compared to other centers, but even such a fragment allows you to identify the connections of Warsaw with Odessa and Kiev, i.e. between the largest regional centers in the west and southwest.

Flows along the Volga retained their importance in 1878, despite the development of railways. There are no separate data on settlements at the Nizhny Novgorod Fair in the reports of the State Bank, but the graph shows payments from Nizhny Novgorod to Samara and Saratov, which may be a consequence of the delivery of goods to the fair. The same category includes cash flows from Yaroslavl to Samara, which means that, as already noted, goods were going up the Volga towards St. Petersburg. Connections with the Volga (Nizhny Novgorod) and simultaneously with the Black Sea (Odessa) Rostov-on-Don and Tiflis had basins in the south, and the commodity flow from Odessa to Tiflis was 9.5 million rubles. it was the second largest in 1878 . Figure 4. Transfers between branches of the State Bank in 1878, with the exception of St. Petersburg and Moscow, in the European part cash flows of more than 1 million rubles, in the Asian part - more than 0.5 million rubles.

* — Data on transfers from the branches of the State Bank to the Polish Bank in Warsaw.
** — the maps do not reflect the branch of the State Bank in Tashkent.Changes occurred with the export of goods to Arkhangelsk along the rivers of the White Sea basin.

The main flow to the north in 1868 through the Vyatka province was about 1 million rubles, but in 1878 it was reduced to 336 thousand rubles and therefore did not fall into the graph. The reasons for these changes need to be studied. The northern river route from Vologda to Arkhangelsk, on the contrary, increased from 514 thousand rubles in 1868 to 1.3 million rubles, despite the fact that the railway to Vologda from the south has already approached.

The right part of Figure 4 is devoted to Siberia, where in 1878 there was not a single connection of more than 1 million rubles. Large flows in the Asian part of the empire were smaller than in the European part, both in 1868 and in 1878. However, at the end of the 1860s, a large cash flow from Tomsk to Irkutsk was noticeable in Siberia, corresponding to the commodity flow in the opposite direction. In 1868, these payments even exceeded the threshold of 1 million rubles, which is reflected in Figure 3. However, by 1878, this flow dried up and amounted to only 126.1 thousand rubles, which may be explained in the course of further research.

Outside of the Tomsk — Irkutsk axis , the topology of Asian cash flows also changed from 1868 to 1878 . In general, there were more major connections (5 in 1868, 12 in 1878). Commodity and cash flows continued to move along the Siberian-Moscow highway, but the destinations and transshipment of goods changed. If the graph of 1868 can be interpreted as purchases of Tomsk and Irkutsk at the Nizhny Novgorod and Irbit fairs, then in 1878 the node in Yekaterinburg clearly dominates, most likely related to the sale of metallurgy products. Goods were exported to the west — to Nizhny Novgorod and Kazan, to the east — to Irkutsk. Yekaterinburg had significant bidirectional ties with Perm. The node in Irkutsk retained its importance, but the export directions were now different — to Kazan and Perm, and purchases went to Irkutsk from Tomsk and Yekaterinburg. To briefly describe the weaker nodes of this payment network, Tomsk sold to Irkutsk and bought in Yeniseisk, Irbit brought goods from the Volga from Saratov and Samara (apparently they were grain) and sold them to Yeniseisk through the fair.

Thus, by 1878, the connections in European Russia were redistributed taking into account the railway network, and in the country as a whole — strengthened, diversified and often changed location, which is not always possible to explain within our knowledge of domestic trade, especially in Siberia.

The changes over the next 10 years are shown in Figure 5 — payment flows of 1888 For the European part of the empire, flows of more than 1 million rubles are shown, for the Asian part — more than 0.5 million rubles. The graph includes 43 connections out of 4,260, or 28.1% in monetary terms of all regional connections. A relatively small proportion of large flows compared to 1868 and 1878 indicates the densification of the network in the sense that small flows have grown over 10 years, but large ones have not intensified, or even weakened, which is apparently an important feature of the 1880s, as discussed below. Figure 5. Transfers between branches of the State Bank in 1888, with the exception of St. Petersburg and Moscow, in the European part cash flows of more than 1 million rubles, in the Asian part - more than 0.5 million rubles.

* — From Odessa to Nikolaev 1.8 million rubles, back 1.2 million rubles.The number of flows in graphs 1878 and 1888 is the same — 43, but the topology of the two payment networks differs in details.

There were no major changes in the railway network during this decade, although it cannot be said that there were none at all. But the economic conjuncture of 1878 and 1888 was different. In the second half of the 1870s, the general situation was unstable, but 1878 was relatively calm. 1888 was clearly worse due to the prolonged depression of the 1880s. Perhaps that is why economic ties were "shredded" at the end of the 1880s. This is noticeable on the largest flows of more than 3 million rubles. In 1878 there were 5 of them with a spread from 4.9 to 13.5 million rubles, but in 1888 out of only 3, and their scale is already somewhat smaller (Tiflis — Odessa, 5.9 million rubles; Riga — Orel, 5.0 million rubles; Odessa — Kiev, 3.4 million rubles).

The node around Odessa in 1888, in fact, remained the same as in 1878 (there were 9 connections became 10), although some of the largest connections were lost, the list of cities changed a little, but in general these are still the same southern regions of the European part of the country. At the same time, the nodes around Kiev, Kharkov, Rostov-on-Don have intensified, perhaps partly due to the fact that in the 1880s there were more railways in the south than in the 1870s.

By 1888, there is already a complete set of data on Polish branches. The main growth of railways in the European part of the empire in 1878-1888 was in the direction of Poland. By 1888, the situation here had worsened in the sense that the Warsaw—Odessa connection was not among the major ones. At the same time, Poland's ties with the south of Russia have been preserved in a slightly different form — from Kiev to Warsaw, from Kharkov to Lodz, from Warsaw to Yekaterinoslav.

Agricultural export along the Riga— Tsaritsyn highway in 1888 retained great importance, which reflected cash flows to Vitebsk, Orel, Borisoglebsk. Voronezh also had access to the Riga direction at that time. However, the export from Kursk to the Baltic now went along the more southern Libava—Romny highway.

In the Caucasus, by 1888, Baku and Batum were connected by railway (via Tiflis), i.e. Baku oil fields received access to the Black Sea. At the end of the 1880s, as before, Tiflis bought a lot in Odessa, but now a large flow from Kiev has also become noticeable. Supplies of Baku oil were being delivered to Odessa via Batumi, apparently. We bought goods for Baku in Kiev.

By 1888, the Volga region's connections were represented by a junction in Saratov, where waterways converged with railways, apparently, grain was redistributed to Kharkov, Rostov-on-Don and Nizhny Novgorod. A large regional cash flow is noticeable from Samara to Uralsk (delivery of goods to Samara). However, in general, Volga connections declined, which could be a consequence of both the development of railways and the depression, partly related to the crisis of falling agricultural prices.

To the description of commodity-money flows in the European part of the country, it is worth adding that the export along the northern rivers, which was discussed above (from Vyatka and Vologda to Arkhangelsk), decreased so much that it did not fall into the graph of 1888. From Vyatka in 1888, the largest flow of goods was already going to Yekaterinburg.

The right part of the graph for 1888 in Figure 5 is devoted to the Asian part of the empire. This segment is noticeably weaker than a similar one in 1878. There are fewer large interregional flows, which can be explained by the general depression, although perhaps this is not the only reason for such changes. At the same time, in 1888 there were flows of more than 1 million rubles in the Urals and Siberia (from Yekaterinburg to Perm, from Irkutsk to Yekaterinburg), whereas in 1878 there were none at all. In 1888, there was a concentration of the main flows between Siberia and Europe through the Urals, which corresponded to the transportation of goods from Perm to Yekaterinburg (via the Ural Railway) and further to Irkutsk along the Siberian-Moscow highway. In addition, in 1888, commodity flows from Yekaterinburg to Nizhny Novgorod (to the fair), as well as to Kharkov, the center of mechanical engineering, were preserved. Apparently, these were sales of Ural metallurgical products. The Irbit junction (fair department) had lost its former significance by 1888. In Siberia, there was a stable interregional payment network, through which goods were transported from Tomsk to Irkutsk, and large volumes of bilateral exchange were maintained between Tomsk and Yenisei. From new, but understandable connections — goods were purchased for Tashkent (Central Asia) at the Nizhny Novgorod Fair.

Thus, by 1888, ties, on the one hand, were losing in volume due to the general unfavorable economic situation, but, on the other hand, this situation cannot be unequivocally called a regression, because development continued through the growth of small ties and the concentration of the most important large ones.

Figures 6 and 7 visualize the payment flows of 1898 separately for the European and Asian parts of the country. The graphs include 120 connections out of 9,284, which in monetary terms was 35.3% of all regional connections in 1898 (more than in the depression of 1888, when there were 28.1% of them). Even a cursory glance at the maps allows us to conclude that the situation has changed a lot in 10 years. Let's try to understand this in more detail.

Figure 6 shows the connections of the centers in the European part of the country (with a volume of payments of more than 1 million rubles per year). The points belonging to the Asian part — Yekaterinburg, Chelyabinsk, Orenburg, Askhabad, etc., but which got on the map in this projection, are represented only by connections with European Russia.

Unlike the previous time slices, in 1898, first of all, the connections between neighboring or nearest centers are noticeable, through which the boundaries of regional markets are visualized. In other words, if in the column of 1888 "distant" interregional connections are noticeable, for example, Riga — Orel, then in 1898 the emphasis shifted to "near" connections around large centers. There is no doubt that this change in network topology was partly influenced by the growth of local railways, which took place in the 1890s.

At the same time, interregional ties are also noticeable on the graph (between Odessa, Kiev, Warsaw, Riga, Kharkiv, Rostov-on-Don, Baku, Tiflis, Nizhny Novgorod). The regional markets around these cities (except Tiflis) are visualized separately and analyzed in our article on the leading financial centers of the Russian Empire ^[10]. Now these markets are depicted on the same map and other links of the same size have been added to them. However, even from this perspective, the strongest and most numerous connections are still concentrated in the west, southwest and south. The Rostov—on—Don Rostov—on—Don network joins the Odessa-Kiev-Warsaw triangle: Riga network adjoins from the north through Warsaw, Kharkiv network from the east through Kiev, Rostov-on-Don network from the east through Odessa, Tiflis and Baku networks from the southeast through Odessa and Rostov-on—Don. In the latter case, quite "distant" interregional ties are observed in the Baku hub, for example, the flow of goods from Baku to Lodz (Poland), apparently it was the sale of products from Baku oil fields. An example of Baku's geographically determined "long-distance" ties is the transportation of goods from Nizhny Novgorod via the waterways of the Volga basin.

The network of Nizhny Novgorod looks weaker than the western and southern regional segments, but it was through this node that connections with the European north (Vyatka, Perm) and the Asian part of the country were provided. Large commodity-money flows along the rivers of the Volga basin are also observed in Rybinsk (from Tver and Novgorod) and in Samara (connections with Ufa and Orenburg). The cash flow from Arkhangelsk to Vologda, which did not exceed the threshold of 1 million rubles in 1888, has now increased so much that it entered the graph of 1898, and in addition to the river routes between Vologda and Arkhangelsk, a railway was already operating.

The connections of the Central Chernozem region with the Baltic ports were reduced and therefore did not fall into the graph of 1898 . The flow between Riga and Vitebsk (on the Riga-Oryol highway) looks more like an intraregional one for the network segment around Riga. Eagle's connections in the west have moved from Riga to Vilna, i.e. to the railways south and west of the Riga—Tsaritsyn highway. Kursk and Voronezh were reoriented from the export along this highway to routes located to the south and west, leading to Kiev and Kharkov.

Figure 6. Transfers between branches of the State Bank in the European part of the Russian Empire in 1898, with the exception of St. Petersburg and Moscow, cash flows of more than 1 million rubles.Transfers related to the branches of the State Bank in Orel and Yelets, i.e. in the Orel province, are analyzed in our article on commodity-money flows of this agrarian region, and partly for the same time slices ^{[12] [11]}.

There is also a sharp increase in "close", intra-regional ties in 1898, although none of them exceeded 1 million rubles. We are talking about the connections of Orel and Yelets with Voronezh, Kaluga, Kozlov, Kursk, Mogilev, Ryazan, Smolensk, Tambov, Tula. This segment of the network was organized in the same way as the segments in Figure 6. In the previous decade, in 1888, the volume of transfers between Orel and Yelets and neighboring cities was no more than 100 thousand rubles. These are very insignificant figures compared to the connections with Riga or Libava ("far", Baltic direction), which amounted to 1 million rubles. a year or more. In 1898, "close" connections are not only more than 100 thousand rubles, but in many cases more than 500 thousand rubles.

Thus, the difference between the graph of 1898 and the previous decades is the unprecedented growth of regional markets, which clearly correlated with the growth of local railways, but this phenomenon needs further study. This process was most pronounced along the western, southwestern and southern borders of the European part of the Russian Empire. Smaller regional markets are visible along the waterways of the Volga and White Sea basins. At the same time, there is a decline in flows along the Riga—Tsaritsyn highway and the redirection of commodity and cash flows to other railways located to the south and west.

The branches of the State Bank in the Asian part of the empire are shown in Figure 7. By 1898, there were noticeably more of them than in the previous decade. However , the time of the Trans - Siberian Railway in 1898 it has not yet arrived, the railway from the European part has reached Krasnoyarsk, and there was also a branch line between Khabarovsk and Vladivostok. For 1898, as for earlier time slices, the problem of different scale of operations in Europe and Asia persists: if for the European part the cut-off limit of small connections of 1 million rubles provides enough material to judge the overall dynamics of the network, then for Siberia and Asia as a whole it is necessary to add connections of more than 0.5 million rubles.

From the old connections, the Tomsk — Irkutsk axis is noticeable, which is joined from the east by new strong flows in Asian segments between Irkutsk and Chita, as well as Blagoveshchensk and Vladivostok. In the west, Tomsk is a hub for connections with Krasnoyarsk (a branch from Yeniseisk was transferred there in 1897), Nizhny Novgorod in the European part and Semipalatinsk in the Steppe Region (Kazakhstan).

All other network segments in the Asian part can be considered as new. Firstly, these are the relations of Western Siberia and Kazakhstan with each other and with the European part. We are talking about branches in Tobolsk, Tyumen, Omsk, Semipalatinsk and their close ties with Nizhny Novgorod, Irbit, Yekaterinburg, Perm. Secondly, the segment of the network around Vladivostok included cargo delivery to Khabarovsk (by rail, the strongest connection), transportation from Irkutsk and Blagoveshchensk, as well as a well-explained direction of sea transportation from Odessa to Vladivostok. Thirdly, in the 1890s, a network was formed in Central Asia - a triangle of strong Tashkent—Kokand—Samarkand ties, as well as export points in Samarkand, Bukhara and Askhabad by rail to the Caspian Sea. With such an arrangement of communications, the region's close contacts with Baku are understandable. The graph also includes two major "long—distance" connections of Central Asia with the European part of the empire - the export of goods to Lodz (apparently cotton) and the import of goods from Kiev. However, despite the fact that the map of Asian flows was greatly updated in the 1890s, and many new interregional ties appeared, the process of economic integration of the region was in motion, because important railways appeared here after 1898.

As a result, the development of regional ties can be traced through four time slices. 1868 is actually a "pre—railway" period when non-capital cash flows are largely a consequence of the transportation of goods by river and horse-drawn transport. The connections of 1878 reflect the formation of the all-Russian railway network. By 1888, these ties had become more complicated and concentrated, although many of them had "shrunk" due to the unfavorable economic situation. In 1898, a new phenomenon was observed — the rapid development of intraregional relations. Figure 7. Transfers between branches of the State Bank in the Asian part of the Russian Empire in 1898, with the exception of transfers related to St. Petersburg and Moscow, cash flows of more than 0.5 million rubles.

Conclusion

The study of the internal commodity-money flows of the Russian Empire helps to clarify the status and system of relations of individual regions in the economic space of the country.

This approach allows us to take a fresh look at the economic diversity of the country, because the processes in a particular region that ultimately affected the well-being of the population could not coincide with the all-Russian trend. However, the study of domestic markets is difficult due to the insufficient number of sources. Statistics of commercial transfers between branches of the State Bank fills this gap, and these are regular comparable data for almost 40 years, in the second half of the XIX century, and it is a pity that their publication stopped at the beginning of the XX century. As it has been noted more than once, data on money transfers are the result of a counter-directional flow of goods and services. The State Bank's monetary relations are the best indicator of internal commodity and cash flows, because in the second half of the XIX century, the State Bank was not only the largest commercial bank serving entrepreneurs, but also had the most extensive network of branches. The processing of these data was carried out by methods of network and geoinformation analysis for four time slices belonging to different economic periods (1868, 1878, 1888 and 1898), which made it possible to assess changes in economic relations over a thirty-year period.

The national payment network had a two-tier structure, in which the capital (connected with St. Petersburg and Moscow) and regional segments were distinguished. The capital's payment network served nationwide markets where regions satisfied a significant part of their demand for goods and services. That is why the capital's markets were very large, they always accounted for more than half of all settlements. At the same time, the regional segment of the payment network was gradually catching up with the metropolitan one.

This study clarifies our previous conclusion about the specifics of the metropolitan segment. As it turned out, these were not two networks of the "star" type, with centers in St. Petersburg and Moscow, but a "double star", because a huge cash flow was maintained between the capitals, which significantly exceeded any flows in this payment network. The double star functioned as follows. In St. Petersburg, sending funds to the regions significantly exceeded the inflow of funds from the regions. This could be because goods arrived in the capital and went further for export from there, or funds received in various ways in the capital were sent to the regions. A significant part of goods and services from St. Petersburg were redistributed to Moscow. In turn, Moscow was a center where the incoming and outgoing flows associated with the regions were somewhat smaller and more balanced compared to St. Petersburg. At the end of the 1890s, the incoming flow prevailed in Moscow, corresponding to the export of goods from Moscow. In this sense, Moscow was similar to the leading regional centers (for example, Odessa, Kiev, Warsaw), but significantly exceeded them in terms of operations, which indicated a higher status of Moscow relative to these centers.

It is important to take into account the growth of regional, non—capital connections when assessing the dynamics of the payment network, because in 1868 there were 745 such connections, or 17.4% in monetary terms, and in 1898 - 9,284, or 43.7%. This dynamic reflected the gradual complication of regional markets, as well as their institutionalization, i.e. the transition of an increasing number of payments from the informal to the banking sector.

For the European part of the empire, the location of regional commodity-money flows in 1868 indicates that the "pre-railway" period has not yet ended, and the observed flows are largely the result of water and horse-drawn transport. In the 1870s, the topology of the payment network changed due to the appearance of major railways. "Long—distance" regional connections (between remote regions) are of great importance, for example, the flow of goods from Orel to Riga to the Baltic ports along the Riga-Tsaritsyn highway. In 1888, the economic depression led to the compression of many large flows, although the number of small flows increased significantly, i.e. development continued anyway. Railways have grown in the south and west of the European part during this decade, which predetermined the displacement of flows from the Riga—Tsaritsyn highway in the southern and western directions. In 1898, when the period of growth of the local railway network began, "near" connections between neighboring provinces began to grow. This phenomenon needs further study, but, at a minimum, it indicates the intensification and institutionalization (transition to the banking sector) of local transactions. Leading financial centers played a special role in regional exchange, the strongest of which were located in the west, southwest and south of the European part of the empire (Odessa, Warsaw, Kiev, Riga, Kharkiv, Rostov-on-Don, Tiflis, Baku).

In the Asian part of the country, there have always been fewer large commodity-money flows compared to the European part. Due to the long distances here, almost all connections are "long-distance", interregional, naturally gravitating to the main tracts and river routes. The stable axis of payment networks in Asia was the Tomsk—Irkutsk axis. Relations with Europe functioned through Nizhny Novgorod (fair). The Yekaterinburg junction was strengthened only in the 1870s. Cash flows in Siberia noticeably shrank in the 1880s. In the 1890s, segments of the payment network appeared in the Far East, the Steppe Region (Kazakhstan), and Central Asia. At the same time, it is obvious that at the end of the 1890s, this network was still only at the beginning of its formation.

Unfortunately, this study cannot be continued into the 20th century, because similar statistics for a later period have not yet been found. However, even the available data show how, in the most general form, the process of integration of the economic space of the empire is underway: first, metropolitan connections are established as the basis for interaction between regions, then, as railways develop, long-distance regional flows are formed, which are redistributed as changes accumulate both in the transport system and in the overall economic situation. Then comes the time to institutionalize close-distance connections. Through these changes, the basic trends in regional development affecting the well-being of the population are noticeable, so it is worth taking a closer look at them.