Railroad investment and regional disparity: Public expenditure on transport infrastructure in France,1837–57

Introduction

Transport infrastructures are generally desirable, but their financing is not simple. Although they necessarily involve public interests and therefore require public intervention, the limited public financial resources and the quest for certain economic liberty or efficiency can lead to private participation. At the same time, as transport infrastructures policies are also an essential part of regional policies, they inevitably concern regional interests. How does the issue of public/private financing interact with this regional dimension in transport policies? We explore this question through the case of investment in main railroad lines in France between 1837 and 1857.

Regarding the issue of public/private financing, the situation of France in the field of public works at the beginning of the nineteenth century was very different from Great Britain's private concession system and insufficient central planning, or from the practice of mixed companies and independent infrastructure provision in various German states. ¹ In France, the central government's intervention in public works was consistent and significant. The Corps des Ponts et Chaussées, a state entity created in the early eighteenth century, was essential in shaping this situation. It was responsible for the construction and maintenance of highways and the improvement of waterways under the Ancien Régime, with growing financial resources. ² This institution recruited from the École des Ponts et Chaussées and the École Polytechnique, two engineering schools founded to select and train qualified engineers for the State, leading to a tendency towards “technocracy” in large public works. ³ The system of the Ponts et Chaussées, organised under an administrative hierarchy, excluded “practically any recourse to private initiatives, in which the role of companies is limited to execution (of works)” and allowed for the uniformity of administrative decision-making processes and the coherence of public works policies in France. ⁴ It survived the Revolution and even managed to expand thereafter. Through the Becquey Plan, an ambitious canal project aiming to build a dozen of long-uncompleted canals thanks to public loans contracted between 1821 and 1822, the Ponts et Chaussées established its domination in the field of canals, which had often been carried out by concessionaires under the Ancien Régime. It was in this context that the rise of railroads occurred. Railway lines, which had started off as private local lines in the 1820s and early 1830s, have finally caught the attention of this institution from 1833, which then tried to extend its dominant role to this new field of public works. ⁵

Regarding the regional dimension, the construction of railroads necessarily interacted with existing transport networks. Before the French State started investing in railroad construction, there had already been evident regional disparities in the distribution of existing transport networks, especially highways (routes royales) and canals. The co-existence of a well-equipped North and a poorly served South was the general situation of geographic distribution of French transport infrastructures. ⁶ Railroads were developed in this context. Should these industrial or commercial centres, which were largely in the northern half of France and already enjoyed dense transport networks, be connected first, or should priority be given to the southern deprived regions in the most urgent need of transport service? Would these investments worsen existing regional inequality? Of course, this delicate situation also existed in countries where railroads were planned regionally or initiated by private actors. For example, in German states, instead of promoting new connections, railroads were first constructed by city-based groups in order to reinforce existing routes, ⁷ and therefore existing transport hubs, with this new form of transport. However, in the French case, the State's desire to extend its central role in the planning and funding of large transport projects to the field of railroads made this geographic choice even more difficult: with a limited public budget, the State had to select which regions would be the first to be equipped with this promising new infrastructure, meaning that railroads would be financed at least partly by all taxpayers, but would only prioritise certain regions.

The relationship between the issue of public/private financing and the regional dimension was therefore important under the July Monarchy (1830–48), when the first general railway plan, also called the “Charter of French railroads” ⁸ , which served as the basis for the French star-shaped railroad network, was approved by Parliament in 1842 and implemented. Of course, the interaction between these two dimensions is not a new topic in French railroad research. Lefranc's study has already shown that deputies were reluctant to vote in Parliament for railway projects serving other departments and financed by the state, thus by all taxpayers. ⁹ However, his study does not address the divergences of regional interests, nor the issue of regional balancing during the vote of the railroad bill of 1842 and its subsequent implementation. These are exactly what we propose to study in this article.

This article aims to explore the important, yet understudied relationship between the public/private financing dimension and the regional dimension at three levels. First, by analysing parliamentary debates in 1838 and 1842 – two critical events for the classification of French main lines and the determination of their financing model, we will show the connection between these two dimensions and its weight in the drawing up of a general plan in 1842. Then, using documents of Chambers of commerce, we will explain how the varying progress of railroad development in different regions impacted the attitude of local stakeholders towards private participation in railroads and how it related to different local interests based partly on the previous distribution of transport infrastructures. Finally, using data from the annual budgets of the Ministry of Public Works between 1837 and 1857, we will measure the regional disparities in transport investments made by the state, and examine the effective influence of railroad financing models on the regional imbalance in fiscal sense during the implementation of the plan of main lines, whose financing model was evolved from a system based on heavy state participation to a private actor dominated one (Figure 1).

OPEN IN VIEWER

Figure 1.

Completed railroad lines in France between 1850 and 1860.

Double inequality in the publicly-funded railroad system: Main lines classification and the introduction of private capital

Regional development theories help us to better understand the relationship between the financial model of transport projects and their influence on regional equality. There are generally two kinds of regional development policies: regionally balanced investment policies and regionally imbalanced ones. The latter are considered by some economists, like Albert Hirschman in the 1950s, as a response to limited financial resources, or even as an effective way to stimulate growth. ¹⁰ The theory of “growth poles” developed by French economists during the 1950s and 1960s also helped to justify imbalanced investments by underlining the positive externalities generated by the “poles” for the rest of the area. ¹¹ As a result of the adoption of a regionally imbalanced strategy, certain areas (natural poles) that are already more developed than the rest of their region are often prioritised to be equipped first by new transport infrastructures. Here arises an important aspect of the relationship between public/private and regional dimensions: such an imbalance project, if financed entirely or partly by the State, can lead to regional inequality in both the territorial and financial senses. This is due to the fact that it is financed by taxpayers from the entire country, rather than by the local taxpayers who actually enjoy the infrastructure. This concept of double inequality is crucial to our later analysis.

To discuss the regional balancing issue in French railroad planning, three regional aspects of transport infrastructures must be considered. First, the weight of local voices in central policymaking increased over the period. With the growing power of Parliament in matters of public finance since the Restoration (1814–30) ¹² and the increasing importance of local stakeholders in transport policymaking from 1829, ¹³ the Ponts et Chaussées's monopoly in defining the utility or public interest of transport projects was partly counterbalanced by considerations of local interest, particularly as advocated by deputies for their department in Parliament and by Chambers of commerce in local inquiries. Second, there were conflicts between financial reality and the local expectation regarding the public exploitation of transport infrastructures. Despite provincial capitalists’ appeal for a state-owned railroad system in order to avoid private transport monopolies, the restrictions of public finances and the much-criticised low efficiency of the Ponts et Chaussées made the private participation in a nationwide railroad project not only reasonable but even necessary. ¹⁴ Moreover, our study about the problem of regional balancing in railroad policymaking should focus on the distribution of all transport infrastructures, rather than the sole railroads. Indeed, in addition to the regional inequalities in the distribution of transport infrastructures before railroad development, there was also the issue of the compatibility between railroads and other transport means. In some regions, the compatibility between railroad construction and other transport projects, particularly canals ¹⁵ , was controversial, because railroads would not only crowd other projects out of limited public or private investment, but also compete with them for traffic.

Let us now turn to the parliamentary debates regarding railroad classification in 1838 and 1842. One key question that was discussed intensely in these debates was: how many main railroad lines (grandes lignes de chemins de fer) should be classified and completed first. This question is directly connected to the issue of regional balancing and the choice of a financial model.

Generally, the proponents of building a single major line were often statists (étatistes) who demanded heavy public investment in railroads and were very cautious about the participation of private actors. For them, it was therefore better to concentrate public resources on the most important line, usually the one connecting Lille or Le Havre to Marseille via Paris and Lyon, mainly for reasons of budgetary limitations. This preference of statists for a single main line is not unique. In Belgium, where railways were centrally planned and initially financed by the State, the government had also introduced plans for single railroad lines, such as that for Antwerp–Liège, in 1832. ¹⁶ Indeed, expenditure on state-financed railroads must respect the limit of the public budget and the State's ability to borrow, otherwise the railroad plan risks being seriously delayed or even suspended. Some deputies refer to the case of aforementioned Becquey Plan, which aimed to build at the same time twelve canals at the charge of the state with public loans, but ultimately suffered major delays, partly due to the insufficiency of funds to cover its significant cost overrun. ¹⁷ This is also the reason why the railroad bill of 1838 was rejected by Parliament: this bill, which required 207 million francs [equivalent to 18 per cent of the fiscal revenue of the French government in 1838 ¹⁸ ] to build four main lines (from Paris to Belgium, Rouen, Orléans and from Marseille to Avignon), was based entirely on the public budget. After the bill of 1838 was rejected, it was clear that, to get a railroad project approved by Parliament, it would be necessary to choose between the development of a single main lines financed solely by the State and the construction of several main lines with the help of private investment.

Although building a single main line owned by the state was justified by the government's budgetary limitations, it would have one severe consequence: the neglect of the majority of departments until the completion of this single line. Such a statist project posed the problem of a double inequality in terms of both territorial equipment and fiscal distribution. This was one of the reasons that convinced the government to adopt the general participation of private capital in railroads. The railroad bill of 1842, which proposed to build nine main lines, a project even more ambitious than the bill of 1838, was thus based on a mixed system, in which private actors were responsible for the construction of railways, leaving public actors in charge of acquiring the land, carrying out the earthworks, building the stations, etc. The general principle for the classification of main lines under the bill of 1842 was to connect Paris to the “land and sea borders, Belgium, Germany, Switzerland, the Mediterranean, Spain, the ocean and the Channel”. ¹⁹ In the final adopted version, there were three exceptions to this principle among the nine main lines: the lines from Paris to Central France via Bourges, from the Mediterranean to the Rhine via Lyon and from the Mediterranean to the Atlantic via Toulouse.

The original bill of 1842 still received two kinds of criticism from deputies. One came from the supporters of the construction of a single line, mainly statists and a few deputies representing the departments that would very likely be connected by the single railroad line. Thiers, the deputy of the Bouches-du-Rhône department ²⁰ in the Chamber of Deputies, observed that, given the limited and already largely committed public budget, “undertaking a grand work, finishing it, and undertaking other works after it is finished, that is what the simplest common sense commands.” ²¹ He assumed that this single line should be the one connecting Lille to Marseille via Paris and Lyon, taking into account its great value to connect economic centres and to resolve “the third concern of the government for ten years”, namely the occupation of Africa. Deputies like him, would like to see the public funds concentrated on the single main line that best served the country's general interest and their region, rather than divided between several other lines. However, such a plan was against the interest of departments not served by the line in question, as it was double unequal towards them. As Alphonse de Lamartine, the deputy of Saône-et-Loire department, argued in Parliament: “By voting for all of these areas of France to participate in proportion to the benefit of the railroads, as they participate in its expenses, I therefore do not believe being unfair either towards the taxpayers who also pay the tax, nor towards the departments, nor towards the State. And by voting for several lines, I am voting for the truth; because I am voting for the geographical configuration and for the equal distribution of the country's taxes.” ²²

The second kind of attack against the bill of 1842 had to do with regional inequality, as many departments would still not be connected by these nine railroads. A deputy of Indre, a centrally located department, expressed his dissatisfaction with the original plan of 1842 which ignored central departments’ need to be connected to Paris and the Pyrenees. According to him, among the 86 French departments, the 20 departments that stood to benefit directly from the bill of 1842 were already served by 9,345 km of highways and 1,491 km of canals, while the other 66 departments only had 7,036 and 233 km, respectively. To make this imbalance even more unfair, the taxpayers of those “neglected departments”, especially the ones in Central France, still need to pay their taxes “to make travel at 10 leagues (1 league≈4 km) an hour the eastern and western taxpayers, to have their goods transported at 48 centimes per ton”, while they could only travel at only 2 leagues an hour themselves, paying 1 franc instead of 48 centimes for the carriage of their goods. ²³ His concerns about double regional inequality were obvious.

Facing such criticisms, the Minister of the Interior justified the railroad bill of 1842 by emphasising the superiority of general interest over particular ones in decisions made by a central authority:

“It is clear that whenever a central power, whether national or departmental, approves an expenditure, it is not certain that this expense will benefit all communities equally; there are sometimes interests that suffer where other interests prosper. But when it comes to the general interest, [when] the expense is good, useful, productive, and [then] we do well to vote for it.” ²⁴

However, the law of 19 July 1845 abrogated this important provision. ²⁶ Figure 2 confirms that the sum directly invested by local administrations was indeed marginal. The preventive measure in the law of 1842 to reduce regional fiscal inequality in public railroad investments, therefore did not serve its original purpose.

OPEN IN VIEWER

Figure 2.

Accumulated capital invested in railroads by companies, the State and local administrations, 1837–57 (francs).

The struggles of local stakeholders to be connected to the railroad network: The trend towards a private concessions system

Although the law of 1842 planned to distribute public expenditures across all of the planned main lines, “the law could only apply funds to sufficiently studied lines”, as recognised by the reporter of the parliamentary commission in charge of examining the bill of 1842, and the precise route of each line had to be fixed according to the difficulty of its construction, as well as the local and general interest it serve. ²⁷ Clearly, it was still the State's territorial priorities, and the profitability or feasibility of each possible route, that would determine the order of construction of these lines and the areas they crossed. We can thus say, the railroad law of 1842 was rather a principle or a general plan than a detailed program; rather a new beginning of a regional competition for public investment than a final result of such a competition. In this section, we will see how different regional interests in this competition affected the attitude of local stakeholders towards the public or private financing of railroads.

Regarding regional reactions to this competition, Yves Leclercq has already presented the financial or non-financial assistance schemes offered by cities or departments to attract railroad companies and to encourage railway construction after 1842. ²⁸ However, despite the dynamic efforts made by local administrations, the scale of financial assistance they provided was still considerably less than the State's railroad investment, as shown in Figure 2. Besides, the power to authorise the establishment of a railway remained reserved to the State until 1865, when this power was finally handed over to the prefects with regard to local secondary lines. ²⁹ Thus, during the period studied here, the most efficient action that local stakeholders could take remained to influence the State's railway policies.

In this subject, the influence of Chambers of commerce must be highlighted. As representatives of local industry and commerce, they have a range of functions in the making and disseminating of State's economic policies, ³⁰ which include transport works policies. Moreover, legislation passed in May 1829 made local inquiry necessary for introducing new transport works projects, ³¹ giving Chambers of commerce a new channel to influence transport policies. Chambers of commerce, especially those of large cities, could therefore influence these policies either by giving their opinion in inquiries, by addressing their advice or complaint to ministers, or through their department's deputies in the legislative process.

The introduction of private capital under the law of 1842 and the trend towards a real concession system in the mid-1840s was very soon confronted with local resistance, particularly from the Chambers of commerce, firm opponents of the execution of the railways by private financial companies and of the tendency towards “privatisation”. ³² However, from the mid-1840s, the different situation of their region no longer allowed them to resist private concessions with the same persistence.

The Chamber of commerce of Lille still advocated that the railroads be financed by the State, even though this would mean that very few lines could be constructed in the near future within the limited public budget. Its position is not surprising, given that Lille would necessarily have priority over railroad service, due to its industrial prominence and its proximity to Paris and the Belgian and English borders. Therefore, its interest lay in focusing public resources on few lines, even at the cost of postponing railroads serving other departments: “By operating more wisely, there would probably be fewer lines built at the same time; but there would also be fewer financial crises to fear, and the State would not be as vulnerable in staying responsible for maintaining and operating these unproductive [rail]roads in the future”. ³³

The situation of Lyon was different from that of Lille. Before the construction of railroads, this city relied on the waterway of the Rhône to reach southern ports and other centres of commerce. The insufficiency of road transport and the constraints of navigation was a problem for Lyon. ³⁴ This made the Paris–Lyon railroad very valuable to the city. Therefore, when the Chamber of commerce of Lyon was informed that without the help of a private concessionaire, the Paris–Lyon line would be forced to stop temporarily at Chalon-sur-Saône, a city at North of Lyon, it finally decided to abandon its statist position and asked the Ministry of Public Works to ensure “the prompt and complete execution of the railroad from Paris to Lyon, whatever the system adopted for its construction and operation”. The possibility that this artery might not serve Lyon right away was clearly seen as being more harmful to its interests than the development of this line by a private concessionaire. ³⁵

The situation in the South-west was even more delicate. The region had been neglected by public investments in canals and highways under the Restoration, and therefore hoped to be compensated by a public investment on its railroad. According to the complaint made by the Bordeaux Chamber of commerce in 1845 to call for the construction of these railroads:

… the southwest of France was almost completely forgotten in the distribution of public works during the Restoration, while 400 million [francs] of canals were opened in the north, east and south. …. All this vast region (the southwest), speaking of which, only received from [the government of] the Restoration approximately 11 million [francs] for the channelling of two secondary rivers, the Isle and the Tarn! His Majesty's Government has put an end to this deplorable partiality and we owe to it the useful works [of the lateral canal of the Garonne] which are now under way; but whatever their importance, they will still be insufficient to put us on an equal footing with the regions which were favoured, to our detriment, for 15 years from 1814 to 1830. ³⁶

Later, when the Toulouse Chamber of commerce was called upon by the Chamber of Lille to rally other chambers in the fight against private concessions, it decided not to support this proposal, ³⁸ regardless of the fact that almost every major Chamber of commerce in the North had given a positive response to the Chamber of Lille. ³⁹ In contrast with the firm statist position of the northern cities, some south-western cities like Toulouse, gave in to the reality that they were not the priority of public railroad investment. To be connected by rail and to avoid being left further behind by other regions, they had to accept even more intensive private participation in railway development than what was arranged in the law of 1842.

The law of 1842 did stimulate railroad construction in France. Figure 3 shows a sharp increase in the length of railroad lines undertaken by the state in 1842. While a mere 10 departments were served by railroads before 1842, by the end of the July Monarchy, this number had risen to 25. ⁴⁰ However, over two-thirds of departments were still not connected by rail. To make matters even worse, the departments equipped with railroads by the end of the July Monarchy were mostly located in Northern France (Figure 1), which had already enjoyed denser networks of highways and canals than the South. In contrast, railroads in the South-west, the Massif Central and Brittany had barely received any public investment. Apart from Brittany, those regions were poorly served by highway and canal networks. In the mid-1830s, the transport network in the South-west was still rather linear than in the shape of “net”; and the central cities were relatively isolated, lacking sufficient transport means. ⁴¹

OPEN IN VIEWER

Figure 3.

Total length of railroad lines delivered and undertaken (not yet delivered) by the State and companies, respectively, 1837–57.

In regard to the Bordeaux–Sète line, which put the South-west on the rail route connecting the Atlantic and the Mediterranean, it remained on paper at the end of the Second Republic (1848–52), despite the efforts made by the Chambers of commerce of Montpellier, Toulouse and Bordeaux and their cooperation with the deputies of south-western departments of Gironde, Gers and Landes. ⁴² In the early 1852, the Chamber of Toulouse once more deplored this regional inequality to ministers, in the hope of drawing the attention of the President of the Second Republic “to the need to complete the main line that would revitalise the commercial interests of the Pyrenean departments, which are the only ones in France deprived of high-speed transport means”. ⁴³ The construction of this line was not started until 1852 by the Midi Railroad Company, who insisted on taking charge of the lateral canal of the Garonne, a yet uncompleted canal which had been financed by the State from 1838. Like the Bordeaux–Sète railroad, this canal, along with the famous Midi Canal, was built to connect the Atlantic to the Mediterranean, and could thus counterbalance the risk of a transport monopoly by the Midi Railroad Company. If in the mid-1840s, the Chamber of Toulouse was still worried that the cost of the Bordeaux–Sète railroad would crowd out the public funds allocated to the completion of the lateral canal of the Garonne, ⁴⁴ the highly disadvantageous position of the South-west in terms of railroad service, led this region even to agree in 1852 to transfer the exploitation of this canal to the Midi Railroad Company. ⁴⁵ The region's desire to be connected by railroad has even outweighed its concern about a transport monopoly. After the railroad concession was granted in 1852, apart from a subsidy of 30 million francs, the State disengaged itself not only from the responsibility of partly financing the railroad, but also from the expenditure on completing and maintaining the lateral canal of the Garonne.

The railroads serving the Massif Central were faced with a similar situation. Their construction had not started until a concession was granted for them at the beginning of the Second Empire. The Grand Central Railroad Company, founded in 1853, undertook to develop these lines without requesting aid from the Treasury. The original concession contract was however revised by the law of 2 May 1855, in which the French State committed to transferring a subsidy of 72 million francs to this company over nine years to ensure the completion of these railroads. ⁴⁶ For a railroad network which was barely profitable yet expensive to build, financial support from the State was indeed necessary. Even with subsidies, the Grand Central Railroad Company barely managed to win the confidence of investors on the financial market. The company was beset by financial difficulties from 1856, and then absorbed by three other companies in charge of neighbouring railroad networks (Paris–Orléans, Paris–Lyon and Lyon–Mediterranean) in 1857. ⁴⁷

The railroads serving Brittany and the western part of Normandy were finally undertaken in 1855 by the West Railroad Company, established after a merger which had already been planned in 1853 but had been postponed to 1855, precisely in consideration of granting the concession of the railroad network in Brittany and Normandy. Indeed, in 1853, this costly merger would have left the newly established company with no financial resources to construct the unstarted railroads, ⁴⁸ whose estimated cost reached 82 million francs. This enormous cost came with a moderate rate of estimated return of 5.5 per cent, without taking into account the high possibility of cost overrun, which was very common in large public works. The government therefore offered a subsidy of 30 million francs to encourage their construction. ⁴⁹

In 1852, the concession system was established in France for railroad construction and operation. As Yves Leclercq suggests, the Second Empire government exercised the dictatorial powers using decrees to approve nearly all the concessions that did not require financial assistance from the Treasury. As a result, although the average annual investment in railroad construction reached 356 million francs [equivalent to 16 per cent of the average annual fiscal revenue of French government] under the Second Empire, which was much higher than its previous level of 55 million francs during the July Monarchy, the part invested by the government was largely shrank, ⁵⁰ and no longer reached the level attained by the Monarchy between 1845 and 1847. These public expenditures were mainly in the form of subsidies and guarantees of minimum interest rate for railroads’ stocks and bonds and subsidies for lines with low or very uncertain profitability. Thanks to this system, the railroad network finally covered the regions of South-west, Massif Central and Brittany during the 1850s.

The evolution of regional inequality in the public investment in transport infrastructure in 1837–57

In this section, we measure regional disparities at different stages of the financial model evolution. As seen above, it is precisely public investments prioritising certain regions that cause the double injustice in terms of both territorial equipment and fiscal distribution. It is therefore important to study these regional disparities in public expenditure on public works from 1837, when the State made its first financial intervention in the railroad construction, ⁵¹ to 1857, when the main lines classified in 1842 were completed. In this end, it is necessary to study public spending on all means of transport, not only on railroads, first of all because, the inter-regional imbalance in the distribution of other transport infrastructures caused or worsened the injustice experienced by some departments over the railway projects of 1842. Second, given the budgetary limit, we must consider the “crowding out” effect generated by rapidly growing railway expenditure on other transport infrastructure projects, as shown in the case of South-west in the mid-1840s. Indeed, the fast-growing railroad expenditure made by the Ministry of Public Works was accompanied by a slight decrease in spendings on other transport infrastructures from 1840 to 1845 (Figure 4). Public spending on transport infrastructures must thus be considered as a whole to measure the overall impact of public railroad investments on regional disparities in France.

OPEN IN VIEWER

Figure 4.

Evolution of annual spending by the Ministry of Public Works on different transport infrastructures.

We first measure the disparity in public spending allocated annually to each department (Ii) from 1837 to 1857 using the coefficient of variation, a statistical tool to measure the dispersion of a group of numbers:

c = ∑ i = 1 86 ( I i − ¯ I ¯ I ¯ ) 2 = v a r ( I ) I ¯ 2

OPEN IN VIEWER

Figure 5.

Evolution of the coefficient of variation of public expenditure on public works in every department, 1837–57.

Considering that the disparity of “ordinary spending” remains stable throughout the whole period, this evolution of the disparity of total spending is driven mainly by “extraordinary spending”, of which the expenditure on railway projects constitutes the majority from 1844, except for 1856. ⁵² The influence of railroad spending on the disparity of annual public investments in each department is therefore evident.

However, despite the clarity of the coefficient of variation, there are two major flaws in using it to measure regional imbalance in the annual budgets of the Ministry of Public Works. First, yearly data fails to represent the magnitude of public investment received by a department, because the expenditure allocated to a transport project lasts for several years, and disappears after the completion of the project. Thus, what we need to measure is not the disparity among departments in annual spending, but the disparity in the magnitude of public investment received by each of them over a certain period of time. Second, the coefficient of variation does not allow us either to identify the “winners” and “losers” of this regional imbalance of public expenditure, or to judge whether these disparities are necessarily unfair. Considering that neither the population nor the surface of each department is identical, the fact that public infrastructure expenditure is low in certain departments is not necessarily the sign of an injustice, but can result from the simple fact that they are smaller or less populated.

Therefore, to remedy these two flaws of the coefficient of variation, we employ the Theil index, a statistical tool often used in regional inequality studies. We choose the Theil index to include the population factor in our measure of the regional disparity of public investment. The formula used to calculate inter-group disparity ⁵³ is as follows:

T = ∑ ( I i I l n I i I P i P )

We divide the 1837–57 period into four phases: the one before the introduction of the 1842 law (1837–42), from 1842 to the political and economic crisis of 1848, the suspension of railroad construction between 1849 and 1851, and the period from the revival of railroad investment in 1852 until 1857, when the main lines were completed. Considering that all the main railroad lines classified in 1842 were designed to connect different groups of departments, and that the progress of these railroad projects clearly varied depending on the region they served respectively, we choose to examine the imbalance between these regions. We therefore identify nine regions based on the following two criteria: the principle of the classification of main lines according to the “land and sea borders” and the actual order of construction of these railways. ⁵⁴ The intergroup Theil index is calculated in Table 1.

In Table 1, we can observe the evolution of the overall regional inequality level across these periods: a substantial increase in inequality in 1842–48, a continuation of this trend between 1849 and 1851, and then its moderation from 1852 to 1857. This largely corresponds to the evolution of the coefficients of variation as shown in Figure 5. However, the index between 1849 and 1851 is not very relevant to our study. Indeed, this period is marked by a sharp decrease in public railroad investment, which lessens the real significance of the high inequality level. More importantly, this three-year period is much shorter than the others. Far fewer public works projects could be executed over such a short and difficult period, concentrating therefore only on a small part of territory and naturally leading to a higher Theil index. In fact, only certain sections of the nationalised Paris–Lyon line and some other isolated short lines in the Basin of the Seine, such as Versailles-Chartres and Montereau-Tonnerre were constructed by the government during 1849–51. We therefore concentrate our focus on the three periods of 1837–41, 1842–48 and 1852–57.

OPEN IN VIEWER

Table 1.

The inter-group Theil index and its breakdown over the four periods of 1837–41, 1842–48, 1849–51 and 1852–57.

	Number of departments	Pi/P*	1837–41		1842–48		1849–51		1852–57
			Ii/I	Theil Index	Ii/I	Theil Index	Ii/I	Theil Index	Ii/I	Theil Index
Basin of the Seine	16	19.58%	24.26%	0.0520	33.62%	0.1817	35.06%	0.2042	28.28%	0.1039
West	11	16.13%	12.70%	−0.0304	10.44%	−0.0454	11.72%	−0.0375	15.36%	−0.0075
North	5	9.32%	10.00%	0.0071	9.13%	−0.0018	4.70%	−0.0322	5.54%	−0.0288
East	9	10.30%	10.00%	−0.0030	10.61%	0.0032	10.93%	0.0065	9.07%	−0.0115
South-east	9	11.32%	9.65%	−0.0154	8.17%	−0.0267	13.77%	0.0269	8.46%	−0.0247
Mediterranean Arc	10	7.73%	8.94%	0.0129	9.53%	0.0199	5.90%	−0.0160	8.49%	0.0079
South-west	9	8.57%	7.76%	−0.0076	5.16%	−0.0262	3.05%	−0.0315	4.47%	−0.0291
Atlantic	9	10.24%	12.43%	0.0241	9.98%	−0.0026	11.08%	0.0087	10.16%	−0.0008
Centre	7	6.80%	4.26%	−0.0199	3.35%	−0.0237	3.80%	−0.0221	10.16%	0.0408
Inter-group Theil index				0.0197		0.0784		0.1071		0.0502

Note: *Due to the lack of population data for the 1840s, we apply the data from the 1851 census to all the four periods. This is an unfortunate flaw in our data analysis.

Source: BNF, 4-LF255-1, Compte définitif des dépenses de l’exercice 1837 (−1857); Ministère de l’agriculture, du commerce et des travaux publics, Documents statistiques sur les chemins de fer, Paris, Imprimerie impériale, 1856, Série hors classe, Tableau N.25, “population, recensement de 1851”.

In our comparative analysis of the regions, in order to avoid certain effects generated by the “artificial” grouping of departments, especially due to the varying number of departments in each region, we analyse the ratio of the proportion of funds received by each region (Ii/I) to its population share (Pi/P). If the ratio of a region is above 1, it is favoured by public investment; below 1, the region is disadvantaged (Figure 6).

Throughout the three periods, only two regions were consistently favoured by the public investment: the Basin of the Seine and the Mediterranean Arc. The privileged position of the Basin of the Seine was even heavily reinforced between 1842 and 1848, the period when the “mixed” system of 1842 was implemented. These two regions, combined with the East, were the three major beneficiaries of this system. Note that the heavy public investment in the North is clearly justified by its very significant population share, and therefore cannot be understood as unfair favouritism. OPEN IN VIEWER