ABSTRACT The role of finance in economic growth is assessed here as part of a comparison between two European peripheries – Scandinavia and southern Europe- in the second half of the nineteenth century. It reveals that financial development was much greater in the former region, a fact explained because Scandinavian financial institutions were better not only at mobilizing the funds of the public but also at transforming them into credits available to the non-financial private sectors of the economy.
Different per capita income levels, through savings, influenced this outcome but exogenous factors arising out politics, society and culture appear to have been more important still in this process. Financial Systems in the Periphery: A Nineteenth Century Comparison of Scandinavia and Southern Europe Jaime Reis Instituto de Ciencias Sociais Rua Miguel Lupi, 18 r/c 1200 Lisboa Portugal jaime. [email protected] . ul. pt 1. Introduction One of the strongest ideas in circulation concerning economic growth attributes an important role to financial intermediation in determining long run macro economic performance.
Recent comparative studies have shown that the rise in real GDP per capita since 1960 is strongly associated with the size, structure and scope of the respective financial systems (King and Levine, 1993; Levine, 1997). The reasons for this have been clear for some time. Financial institutions provide money in response to the mounting demand for real cash balances and consequently help reduce the cost of transactions. They supply improved entrepreneurship for firms, which enables the latter to use financial resources better and thus to contribute to increases in TFP.
By gathering information on clients, projects and technologies more cheaply, through economies of scale, they become able to reduce risks, attract more savings and combine the latter into loans of a scale that would be beyond the reach of non-institutional actors. Their capacity to monitor borrowers over time permits investment in projects with a higher average return (and risk). All of this implies a larger amount of finance available at a relatively lower price and a better allocation of resources.
Historical studies in this vein, although part of a long tradition, are not frequent and have often been based on single countries. When comparative, they have focused prevalently on the regulatory and institutional aspects of the problem, asking what was the impact of this on the efficiency with which banking systems allocated resources among competing ends. The basic element in these enquiries is the classic dichotomy between two contrasting arrangements: the universal, bank-oriented, Continental system, and the Anglo-Saxon, functionally specialised, market oriented system (Ziegler, 1998).
Although fundamentally concerned with the influence this had on growth, only a few of these studies have actually tested this relationship quantitatively, and the debate over the respective allocative merits of these systems continues to rage, still, it seems, inconclusively (Kennedy,1987; Calomiris, 1995; Collins, 1998; Fohlin, 1999). Some studies, however, have directed their attention to a more rigorous verification of the generally assumed relationship between finance and economic growth.
These exercises have shown that from the mid 19th century long run structural relationships did exist between the two variables. They have also brought to light, in some cases, the exogeneity of the financial variable and the fact that finance Granger-caused growth (Hansson and Jonung, 1997; Wachtel and Rousseau, 1995 and 1998; Rousseau and Sylla, 2001).  Any such research obviously requires a considerable amount of compilation of historical financial statistics.
The bulk of the historical literature has paid relatively little attention, however, to quantifying the evolution of the scale and structure of national financial systems over the long run and making these results truly comparable across borders. This is surprising since from earliest times (Cameron, 1967) it has been known that, in this respect, there were major differences between countries. Moreover, most efforts in this direction have been lopsided and have concentrated on commercial banks alone, rarely considering savings banks or stock markets, and even less the totality of institutional forms that constitute the financial system. 3] In particular, they ignore national contrasts in terms of the efficiency with which these systems ‘transformed’ the savings they collected into funds to be invested in the non-state, non-bank sectors of the economy that really powered the long term growth process. It is these credits that are the crucial link between finance and growth and the purpose of this article is therefore to evaluate how important these differences could have been. The framework for this is the comparison between Scandinavia and southern Europe during the period 1850-1914.
The ‘long second half of the nineteenth century’ considered here is interesting because of the process of convergence and the sustained high rates of growth that were generally experienced in the West. It was also a time of rapid development of modern financial systems across Europe. According to Goldsmith (1985, p. 2), ‘the creation of a modern financial superstructure […] was essentially completed in most developed countries by the end of the 19th century or the eve of World War I’.
In both aspects, however, it was also an epoch during which significant disparities emerged between the two groups of nations that we have selected for this study – Scandinavia (Denmark, Sweden and Norway) and southern Europe (Italy, Portugal and Spain). From being similarly poor around mid-century, their respective growth paths diverged, with the Scandinavian economies achieving growth rates above the average, and their southern counterparts lagging consistently behind the European norm.
By 1913, according to Prados’ (2000) figures, the average GDP per capita of Denmark, Sweden and Norway was 30 per cent higher than that for Italy, Spain and Portugal. A similar contrast is regularly made as regards their financial development. Scandinavian countries have been regarded as ‘financial sophisticates’, Sweden in particular owing to an unusually early abundance of financial institutions and instruments, while the southern European countries have been portrayed as emerging late and slowly in this respect. In what follows, we deal with two issues.
The first is to establish, in as rigorous a way as possible, how large in fact was the financial gap between southern Europe and Scandinavia during the period. Given that national financial systems can be highly idiosyncratic, this means cutting through the variety of forms and appearances in them and focusing on an objective yardstick that is both applicable in all economies and relevant to the problem of long term growth. In accordance with the recent literature, we have adopted the per capita volume of credit supplied by all the components of the financial system to the non-state non-bank sectors.
Having established that this difference was indeed stark, the second aim of the paper is to account for the differences in performance of these systems and in particular for the role of the exogenous causes. We do not seek to measure the actual impact of finance on the growth of these economies, an important task but one for which, as argued above, adequate data is not yet available. In what follows our steps are guided by two concerns. One is that the only proper framework for this study is the financial system as a whole, i. e. omprising therefore every type of institution and instrument, rather than, as often happens, only some. Recognition is hereby given to the essential fact that in different countries similar institutions may not always have performed the same function, just as apparently analogous functions may have been performed by different institutions (Nordvik, 1993; Eichengreen, 1998). The second is that though individual institutions may be formally identical, they can still diverge considerably in terms of their success in playing their designated roles within the economy.
This difference can depend on a variety of factors, including regulatory, political and cultural ones, and even the timing of their development (Guinnane and Henriksen, 1998). The paper is divided into four parts. The first one outlines the size and structure of the financial systems of the six countries considered, and compares the respective degree of ‘financial sophistication’ as measured by their aggregate financial liabilities. These data are presented in per capita terms, at ten yearly intervals, and converted to a common currency (sterling).
The second part analyses, again comparatively, to what extent these resources were transformed into credits and made available to the economy. The crucial distinction here is between those that were supplied to the productive sector and those that were not, because they were absorbed by the state, were retained by the non-state financial system, or employed in foreign operations. Parts three and four advance reasons for the inter-country differences encountered in the preceding two sections and are followed by a conclusion. 2. Financial systems: size and structure
Over the course of the second half of the nineteenth century, a variety of financial agents were active in every European country with the aim of mobilising savings in order to make them available for use by others. These funds were actively competed for by different sorts of institutions whose efforts had a profound influence on the shape of the national financial systems which emerged (Forsyth and Verdier, 2003). Analytically, one should differentiate between broadly three groups of actors, according to their institutional nature, the instruments they used, and the applications they made with these resources.
These are the corporate financial sector, the market for issues by the non-financial, non-state sector of the economy, and the state. The aim of this first part of the paper is to measure the size of each of these sectors in order to gauge, finally, the latter’s overall size. At this point, we are not making any distinction regarding who supplied these funds, in particular whether their origin was national or foreign. The concern is simply to examine the capacity of each national system to persuade the public to accept the claims it issued. This is our first measure of systemic efficiency.
One of the main features of the financial development of Europe during the nineteenth century was the emergence of a myriad of types of corporate institution in this field. These differed as to legal form and ownership, the sorts of guarantees they offered to their depositors and shareholders (if any), the operations they were legally allowed to carry out, and whether or not their object was profit. Table 1 presents data between 1860 and 1910, at ten yearly intervals, on the aggregate liabilities of corporate financial institutions in all the countries in our sample.
Every type of corporate financial institution then in existence is included – joint stock commercial banks, savings banks, co-operative banks, postal savings banks, pledge banks, industrial and artisans’ banks, mortgage banks, credit associations and so on. Our intention here is not to portray the precise evolution of these systems over time but merely to provide a set of benchmarks to enable us to see roughly how they behaved in the course of this period of momentous change. For the sake of comparability between countries, these data are expressed in per apita terms and have been converted into pounds sterling. The latter choice is justified on the grounds that these resources had a high degree of international mobility and that the period was one of exchange rate stability. It would thus seem that the differences between this option and one employing purchasing power parities would not be relevant to the argument presented here.  [table 1 about here] The first thing to note is that by the earliest date (1860), a considerable gap already separated Scandinavia from the southern European countries.
Even though Norway was far behind Denmark and Sweden, its financial institutions still managed to almost double the efforts of Portugal and Italy though not Spain, then undergoing an exceptional banking boom. On this showing, it seems clear that the former might fairly be termed early ‘financial sophisticates’ relative to the latter, particularly if we bear in mind that GDP per capita levels were not so dissimilar.  Neither the passage of time nor the long-term process of economic growth, eroded this advantage, rather the contrary happened.
In terms of group averages, the ratio between the per capita corporate financial liabilitites of the two regions rose from 3 to 5:1 between 1860 and 1910. Relative to GDP instead of population, the southern European countries fell back even further. In this case, the inter-regional gap increased by 100 per cent.  Within the two groups, per capita distances also changed somewhat. Italy significantly outdistanced Spain and Portugal, while Denmark pulled away from Sweden, and Norway converged on both of them.
As in the more industrialised nations, this period also witnessed a strong expansion of stock markets in these six countries and therefore of new ways of mobilising financial resources for productive investment. Even on the periphery of Europe, significant numbers of corporate non-financial firms issued claims against themselves and placed them in the hands of the public through these markets. In 1900, close to 200 such entities were active in Denmark, of which one fifth were banks (Neymark, 1903: 190), while in Sweden, these figures were respectively 165 and 66. 7] In the same year, Portugal had 22 non financial companies quoted on the Lisbon stock exchange (Diario do Governo, 1900) whereas Spain had 54 quoted in Madrid (Boletin de Cotizacion Oficial, 1900). Meanwhile, in Italy, the Milan bourse quoted 57 companies in all (Curioni, 1995).  Consistent and reliable aggregate data on the results of this activity are difficult to obtain and those presented in table 2 must be treated with care. They comprise the total nominal value of the bonds and shares issued by all non-financial corporations quoted on the respective national exchanges.
Issue prices are used to value these stocks owing to the difficulty in gathering current price information for the entire market and period. Mortgage bonds are excluded, to avoid double counting, given that they were included already in the liabilities of all corporate issuers of such paper compiled in table 1.  [table 2 about here] The main conclusion to draw is that the superior capacity evinced by Scandinavian financial corporations to attract savings is revealed here too. On the other hand, the contrast is not as strong as in table 1, and only emerges clearly towards the end of the century or perhaps as late as 1910.
Before that, Sweden and Norway were barely ahead of the southern European economies. Among the Scandinavian countries, only Denmark had outdistanced these followers by as early as 1880. It is also worth noting that in the ‘less sophisticated’ south, bonds and shares helped to mobilize a larger volume of financial resources compared to the banking sector than happened in the north. Until the beginning of the twentieth century, the financial liabilities registered on the stock exchange in Spain and Portugal were almost equal to those of their financial corporations (table 1), while in Italy they were half as large.
In Scandinavia, these proportions lay between one seventh and one half. During this period, the single most important source of claims on the stock exchange, in any European country, was the state, a player which tends to be forgotten in these calculations. Table 3, which makes this abundantly clear, also shows the varying extent to which, in different countries, public debt imstruments were sought after by savers in search of financial applications. In contrast with table 2, the figures presented here are based not on the face values of the bonds in question, but on their current market value.
In the case of Scandinavia, where these instruments were usually close to par, this matters little. The same cannot be said, however, for Europe’s southern rim, where public bond prices were normally subject to hefty discounts at issue, and fluctuated quite strongly thereafter, reflecting lack of investor trust, even when they were denominated in gold and quoted abroad. The result is a fairly strong downward revision of the Italian, Spanish and Portuguese nominal data, throughout the period in the cases of Portugal and Spain, or during its earlier decades in the case of Italy. table 3 about here] One unexpected conclusion is that prior to 1880, despite their reputation for profligacy, the public debt of the southern European countries did not absorb real resources to the same extent as Denmark although it was clearly ahead of Norway and Sweden. After the 1870s, however, the situation changed radically and the activity of the state as financial actor became so intense in these countries that it managed to attract, via the market, a volume of savings that was often two to three times greater than that captured by the corporate financial sector.
In Scandinavia, the opposite happened. After 1870, bank type institutions gathered resources that were altogether several times greater than those the state managed to place with the public. Table 4 puts together the information displayed in tables 1-3 and provides us with a complete picture of the financial systems of our six countries as they evolved over time. What emerges is that although the Scandinavian economies led the southern ones, they were less the ‘financial sophisticates’ that one might have thought judging simply from corporate activity levels. 10] Whilst the north/south ratio of table 1 steadily rose over the second half of the 19th century and reached circa 5:1 by 1910, that of table 4 remained more or less steady at a far lower level, around 2:1. Considering that the corresponding regional GDP per capita ratios were never in excess of 1. 5:1, this seems reasonable. A surprise indeed would have been if we had had to conclude from the data that the Scandinavian economies succeeded in mobilizing four or five times more financial resources and yet only managed in the long run to grow a half percentage point a year faster. table 4 about here] The contents of table 4 suggest three further remarks. The first concerns the vast superiority of the state in southern European in raising funds through the market, which compensated for much of the backwardness of the two other sectors and overall rendered these countries less ‘unsophisticated financially’ than if we had ignored this dimension. This occurred because in Italy, Spain and Portugal the state provided investors with a superior guarantee – the public revenue – and had no limits as to the price it was prepared to offer – not unusually 6 per cent or more.
On the other hand, it confirms the presence of a serious problem of ‘crowding-out’ of the private sector by the public one in these countries.  The second comment is the speculation that the southern European countries’ private intermediaries could have done better had their respective public debts been a good deal closer to the level of their northern counterparts. A simple counterfactual based on the assumption that both banks and the state drew their funds from the same pool of savings – a view that some, e. g. Verdier (1997), would not accept – suggests an interesting result.
In 1910, if Spain had had a Scandinavian scale of public debt and channelled the resulting surplus funds into its corporate financial sector, the latter’s liabilities would have been, ceteris paribus, nearly three times larger than was the case. In Italy, they would have been almost twice the actual size. Such was the price paid in terms of the underdevelopment of corporate banking before World War I for the exceptional financial power of the state. The third reflection has to do with the feedback mechanism that is usually assumed from the growth of an economy to the expansion of its financial system.
On the assumption that this was part of a structural relationship, one might have expected the relatively slower growth of per capita GDP in Italy, Spain and Portugal to be matched by a slower expansion of their financial systems. Yet, table 4 shows the opposite. Over the long run, although they did not converge, neither did fall back relative to the Scandinavians. Consequently, despite a weak start on the financial league table, southern European financial growth exceeded what might have been expected of countries where the economy that drove financial development was advancing more slowly than in the north of Europe. . Recycling liabilities into credits Knowing the size of a country’s financial system tells us about its capacity to marshal resources but is of little help in establishing its contribution to economic growth. A given financial system might have been large in terms of its liabilities but could have given a relatively weak impulse to growth, and conversely. This would depend largely, though not exclusively, on what proportion of the funds gathered was made available, through these intermediaries, to economic agents in the productive sectors of the economy for purposes of capital formation and technological development. 12] This varied considerably among the constituent parts of each system, as well as, in each case, over space and time. One reason is that ‘apparently similar institutions can yield different economic outcomes depending on the context in which they develop’.  In the second place, financial systems are rarely structured in identical fashion. In some countries, savings institutions were more important than in others, where commercial banks prevailed, while in still others it was the central bank that wielded the greatest influence.
Why this diversity occurred depended first and foremost on their individual histories, but not only. They were also shaped by law and politics, by the demand profile of credit users, by the underlying structure of the economy, by culture and by the amount of human capital present in the society. In each country, the proportion of financial liabilities that became credits depended therefore on the structure of its financial system and on the efficiency with which its various components recycled financial resources into credits and made these available.
In order not to overload the exercise, we shall confine our attention here to the situation in 1900 in the six countries under observation and will start with the corporate financial segment. Sweden provides a good illustration of what can be encountered in the corporate sector of Scandinavian countries. Among commercial joint stock banks, the ratio of credits (to the non-bank non-state sectors) to total assets was around 76 % (Statistik Arsbog for Sverige, 1900).
The remaining assets were employed in cash reserves, foreign trade operations, ‘other assets’, including’bricks and mortar’, and bonds and shares, which are excluded from this analysis to avoid double counting. The Riksbank, still a commercial-cum-central bank, could only manage a value of 59 % for this ratio, mainly owing to the higher reserves required by it note issue. On the other hand, mortgage banks supplied 92 % of their liabilities in the form of loans while savings banks reached the 87 % mark.
Such high rates of transformation were possible because in the case of not-for-profit institutions, cash reserves and fixed capital requirements were small while foreign operations were not part of their core business. Investments in bonds and shares came to relatively little too. In the south of Europe, the distribution of these ratios was similar but in each segment the level of financial efficiency was lower. Commercial banks in Spain, for example, lent about 40 % of their assets to the non-state, non-bank sector (Tortella, 1974). In the case of the Bank of Spain, this ratio was a mere 15% in 1900.
This was an exceptional figure due to the Bank’s holding of an unusual amount of Public Debt associated with the Cuban war, but in a ‘normal’ year it lay between 20 and 25 %. Comprehensive data for savings banks and some monographic information indicate that although reserves were small, most available funds were invested in state bonds and railway shares and bonds. At the turn of the century therefore ordinary loans absorbed no more than 30 per cent of total deposits. (Nadal and Sudria, 1983; Martinez Soto, 2000). Finally, the mortgage sector supplied 59% of its assets to the credit market.
Table 5 displays the considerable variety of situations to be found in the six economies under consideration, in 1900, as regards the structure of their respective financial corporate sectors. In Spain and Portugal, joint stock commercial banks (including banks of issue) had a crushing weight compared with the rest, while Sweden, though less pronounced, followed in this group. In Italy and Norway, on the other hand, there was a better balance between profit and not-for-profit institutions, whereas in Denmark the latter, more ‘efficient’ segment was heavily dominant.
The figures in column 1 of table 6 are the result of combining these shares with the corresponding sector’s ‘transformation ratio’ and for the benchmark year they give us the global ‘transformation ratio’ for each country’s corporate sector. The point is that in the late 19th century Scandinavian corporate financial institutions were not only far better at mobilising savings than their southern European counterparts, as we saw in the preceding section. They also succeeded in converting these liabilities into a relatively larger mass of credits to the productive part of the economy to use for investment purposes. table 5 about here] The above is only part of the efficiency story of intermediation in these six financial systems, however. To be complete we still have to incorporate the contributions of their remaining components, namely those created by the issue of credit instruments by the state and by the non-bank corporations. Regarding the flow of funds originated by the latter and for want of better information, it seems fair to assume a 1:1 correspondence between credits supplied to borrowers and the liabilities issued on the stock market, although a slight overestimation may be the result.
The public debt is harder to handle given the difficulty in establishing what part of the revenue from its placement was applied to productive investment. Certainly not all of it was dissipated in useless current expenditure, as the concept of ‘sterilisation’, so frequently used in this context, insinuates (King and Levine, 1993). A part was certainly used to create social overhead capital that would enhance the economy’s output and raise total factor productivity, mainly through rail, road, telegraph and port construction.
Mata (1993: 273) has estimated that in Portugal, perhaps an extreme case of deficit financing, between 1852 and 1914 some two thirds of the resources raised in this way were absorbed by the repayment of pre-existing debt and the financing of current debt. For Spain a comparable proportion is quoted (Comin, 1988: 637 and 652) and we have therefore extended this ratio to all three southern European countries, since Vicarelli (1979:144-5) also suggests for Italy a similar situation.
By contrast, Scandinavian governments borrowed little to finance budget deficits since these were small and infrequent. We have therefore assumed for all of them, conservatively, that two thirds of these funds were directed towards productive investment.  [table 6 about here] Putting all this together brings to light (table 6, cols. 4 and 5) a dramatic contrast in the efficiency with which Scandinavian and southern European systems performed their role as financial intermediaries. The former’s lobal capacity to recycle liabilities into credits was twice as great. Combined with a greater aptitude to mobilize savings, this meant, in absolute per capita terms, the ability to provide the non-state, non-bank sectors of their respective economies with three times the financial resources. Relative to GDP this advantage was almost as substantial, since at this time the Scandinavian economies by this yardstick were ahead by only one third. There were two reasons for this stark difference in recycling capacity.
To begin with, whichever type of financial institution is considered, the Scandinavian ones were always more effective. In the second place, in terms of the structure of the system, those of the southern European countries were all biased towards the type of institution with the lower transformation ratios. They had a greater weight of the public debt relative to corporate finance and the stock exchange, and among corporate institutions, central banks were dominant over the rest, while the highly ‘productive’ non-profit sector was generally less developed than the profit one.
By the turn of the century, the Scandinavian countries had thus become ‘financial sophisticates’ but in a broader sense than has been argued to date. In a growth perspective, no doubt it made some difference that they enjoyed a greater use of monetary instruments, had a higher density of banking outlets and could muster a more substantial volume of deposits. What was decisive, however, was their enormously superior ability to inject financial resources into the real economy and this is something that most of the literature has missed.
Thus, even without a rigorous estimation of the finance- growth relationship for all six economies, it is hard to escape the conclusion that finance could not but have played a significant role in the divergence of growth patterns that distinguished the two peripheries during the 1850-1910 period. The issue this raises is why these two sets of countries moved so far apart in their financial performance and to what extent this was the result of exogenous forces? The second point is of particular interest because it raises the counterfactual question: did these backward economies really have to be so?
The following sections attempts to address some of these problems. 4. Why Scandinavian financial systems mobilised savings better Given its broad sweep, the discussion in the following two sections does not pretend to exhaust the subject. Its aim is to identify the main inter-country differences that shaped the long-term development of these six financial systems. In this section we look at two aspects which appear to have been determinant. One is the volume of resources that were available in these economies to be mobilised.
The other is the propensity of savers to acquire the liabilities issued by their national financial systems The quantity of savings that an economy is able to accumulate is a function, among other things, of its per capita income level and its growth rate (Loayza et al. , 2000). Other things being equal, better-off and more dynamic societies have not only a larger amount of available resources out of which to save but also a higher propensity to do so. Throughout these years, the Scandinavian countries would appear to have been at an advantage in both respects relative to southern Europe.
Before we analyse this relationship, however, we must deal with two potential distortions. The first concerns the probability that not all the financial resources considered in the preceding sections originated domestically. At a time of great international factor mobility, a poorer economy might well have a weaker domestic supply of savings to fuel its financial development, but be able to compensate this by attracting foreign-owned capital. On the other hand, more developed financial systems might reinforce their advantage by drawing in, additionally, large amounts of funds from abroad.
Table 7 shows that at the end of our perio such net inflows were indeed contributing significantly to the growth of all financial systems but far more in the Scandinavian case. The latter’s superiority in attracting domestic resources was thus matched by a similar strength in the international sphere, a fact that has been noted before (Rousseau and Sylla, 2001). On the other hand, this did not alter much our previous ranking of these countries and still leaves to be explained the considerable gap between the two groups of countries in terms of financial liabilities per capita. table 7 about here] A second potential source of distortion was the effect of hoarding on financial activity. Given the alleged inclination of southern Europeans towards this form of storing wealth, as might befit traditional peasant societies, it seems fair to ask how much of the region’s weaker institutional savings performance was due to this. For the sake of argument, we suppose that Scandinavian countries were too advanced, socially and culturally, to engage in such practices.
We further assume that in southern Europe this concealed wealth would have taken the form mainly of gold coin – silver would have been too bulky – and therefore the stock of this type of specie would have been its upper limit. In the late 19th century, this varied between ? 1. 4 and ? 2. 4 per capita, for Italy and Portugal respectively, with Spain somewhere in between (Zamagni, 1993; Reis, 1992; Tortella, 1974). Even if we admit that as much as half of this was hoarded, this would have entailed only a small correction of the figures in table 7, of about ? per capita, thus leaving its essential findings untouched.  Recent evidence based on more than seventy developed and less developed economies, between 1960 and 1995, have established a strong correlation between indicators of private savings and financial development. More importantly, they also provide us with regression estimates that quantify the positive influence of per capita GDP on the savings rate (Beck et al. , 2000). The lack of any comparable evidence for earlier times prevents us from replicating this exercise here but we can follow another, less exact approach.
This rests on the assumption of linearity in the relationship between the two variables, which is suggested by the behaviour of the underdeveloped economies that comprise this sample. Using the data in Prados (2000), we find that average differences, between Scandinavia and southern Europe, of income per capita were respectively 12 % in 1880 and 25% in 1913, and we shall presume that the gap between gross savings must have been similar. The differential in domestic per capita savings absorbed by the financial systems was, however, much greater, respectively of the order of 30 and 100%.
This strongly suggests that besides the effect on financial development of levels of income and savings, other factors, mainly of an exogenous nature, must have played an important part too by causing economic agents to channel different shares of their spare resources to the financial systems of their respective countries. Three features of any financial system are bound to affect the inclination of savers to enter into lasting relationships with its institutions and markets and, in particular, to entrust them with funds.
One is accessibility to users, another is suitability to their specific needs, a third is trust. In what follows we shall concentrate exclusively on the corporate financial sector because, in contrast, differences among countries in the development of their respective stock markets appear relatively minor. In the Scandinavian countries, the volume of private securities was considerably greater than that of state bonds, whereas in Spain, Portugal and Italy the opposite prevailed. Yet when we put all of these financial instruments together, the aggregate volume per capita is relatively uniform throughout the sample.
Moreover, the information regarding national stock markets suggests that inter-country regulatory divergences were not substantial at this time and where they existed, this would not have made much difference to global outcomes (Fohlin, 2002).  Recent research on the post 1960 period has argued that ‘cross country differences in legal and accounting systems help account for differences in financial development’ (Levine et al, 2000: 31) and similar claims have been made in a far broader historical perspective (Sandberg, 1978; Sylla, Tilly and Tortella, 1999).
An overview of the legislation governing corporate financial activity does not suggest, however, that this was a major cause of the divergence we have been examining here. This is not to say that there was absolute regulatory uniformity within the sample or that the legal framework had no impact on other aspects of the financial history of these countries. Rather, there could and was an influence but the effect was not necessarily important in the present context. The Norwegian-Swedish comparison illustrates this.
Regulation restricted the lending policies of savings banks in the latter country while it was quite liberal in the former. As a result, the Norwegian savings sector flourished and even took on the functions of commercial banks. In Sweden, on the other hand, it did only half as well and was overshadowed by the commercial sector, which in Norway was comparatively weak (Egge, 1983; Nordvik, 1993). Globally, however, the two countries achieved very similar results in terms of the assets gathered by the financial corporate sector as a whole, only through different structures.
There are three ways in which the time-path of regulation could have influenced the evolution of commercial banking, yet in all of them a surprising degree of international uniformity is encountered. Barriers to entry is one of them. After a highly restrictive first half of the nineteenth century, which was dominated by specially chartered national privileged banks of issue created to deal with pressing monetary and fiscal problems, in the 1850s and 1860s it became relatively easy to found joint stock commercial banks with limited liability.
This new ease of incorporation opened the system to competition, vastly increased the number of institutions and allowed them the freedom to open branches, which in some countries proliferated and in others not. The second area is that of the limits placed on the scope and type of business banks might undertake. Typically, rules defining lending policies were few and on the whole were quite liberal. In some cases, banks were simply governed by the general law on joint stock companies, while in others they were placed under a specific banking code (Grossman, 2001).
Towards the end of the 19th century, however, and as a consequence of various crises, regulation was tightened in some countries, but it is unclear whether this entailed much change. The evidence is that the enforcement of bank legislation was on the whole lenient, in the spirit of Liberalism, and the institutions dealt with were often allowed to evade it when this was found expedient (Fritz, 1988). The ease with which Norwegian banks overcame the strictures of usury laws is an eloquent illustration of this (Knutsen, 2003). 17] Finally, a fully centralised note issuing regime does not appear to have produced results that were inferior, in terms of financial development, to those where multiple issue was in place. Despite the contemporary belief that commercial banks that had a right of issue enjoyed a business advantage over deposit banks, one finds instances of both regimes both in Scandinavia and southern Europe with no obvious impact on global financial performance.  Mortgage banking based on the issue of bonds, which was for profit in some countries, but not in others, displayed the most significant degree of regulatory variety.
In Spain and Portugal, national monopolies were established from the start, while Italy experimented with regional ones and went over to a national one in 1890. They therefore tended all towards large loans and large denomination bonds. In Scandinavia, Sweden and Denmark were very liberal on this score, but Norway had a state mortgage bank. Table 5 reveals that in the long run, however, this mattered less to the respective shares of this sector than might be expected.
In Denmark, mortgage bonds were very important but Sweden was on the level of the monopolistic countries, while Spain, with a similar set of rules, did frankly worse than all others, including Portugal,. When it comes to non-commercial, small scale, local and often cooperative banking, it is essential to remember first that this was a era in which two quite different types of saver supplied the resources of financial institutions and arguably constituted quite separate segments of the market (Verdier, 1996).
The well-to-do deposited with or bought the shares of commercial banks, while middle and low income people prefered to entrust their savings to local savings banks, credit co-operatives and the like (Vittas, 1997). This being so, in societies where both kinds of institution had a significant presence, as happened in Scandinavia, the financial system was likely to collect a larger portion of savings. In Spain and Portugal, the weakness of the savings sector was such that it meant that a majority of the population in effect had little access to the system as a whole.
Consequently, the volume of savings per capita gathered was smaller. Italy was able to do better than the Iberian peninsula because of a considerable and varied movement of thrift organisations prevalently in the northern half of the country that tapped, the savings of the humble and middle class people, whether rural or urban. It is far from evident, however, that national dissimilarities in legislative framework were responsible for this kind of institution to evolve to such contrasting extents.
In all countries considered, thrift institutions were supported by local or national authorities in a variety of ways, with deposit guarantees and, perhaps most important of all, with tax exemptions. Arguably though, on the assumption that there was market segmentation, this would hardly have diverted funds from the commercial joint stock sector, which was the principal alternative. Until the 1880s, the general norm was absence of legislation, complete ease of entry, an enormous multiplicity of statutory arrangements and only slight restriction on the uses to which savings could be applied.
Full and proper regulation had to await the 1880s – 1875 in Sweden – but, in the event, was of a very mild nature.  Supervision, accounting rules and some operating limits were introduced, against the solid resistance of the thrift institutions themselves, but by most accounts this barely influenced the sector’s level of activity (Bruck et al. , 1995; Hansen, 2001). One of the principal aims of regulation was to stimulate trust in the system by deterring irregularities and imposing transparency on its operations.
As we have just seen, Scandinavia’s superiority in mobilising resources does not seem to have owed much to a better set of rules and regulations. From the public’s point of view, a more reassuring indicator of trustworthiness was how a system performed, not its rules, and here stability of markets and institutions was doubtless the factor that would affect the inflow of savings. Whilst all countries were prone to turbulence and its savers and investors suffered losses as a result, between 1860 and the First World War, the southern European record seems to have been by far theworst.
One instructive sign of this is the mortality of commercial banks. Complete data are only available for Spain, Italy and Denmark but are highly revealing. In Spain, aside from the earlier devastation wrought by the crisis of 1864-6, of the 117 banks founded after 1874, only 60 were still open in 1914 (Tortella, 1974). The losses to Italian commercial banking during three critical periods were similarly substantial: 42 out of the 143 in existence, in 1873-9; 21 out of the 161, in 1888-93; and 11 out of the 163, in 1902-4 (Mattia, 1967).
In Denmark, of the 160 banks created between 1845 and 1914, only 20 failed. A second measure is the variance around the trend of a global systemic indicator such as total assets. Available information covers only the same three countries but the result, now comprehending all types of banks, fully confirms the earlier finding.  The stability of the Danish financial system was significantly greater than that of the Italian one, with the Spanish one a long way behind.
If a stable financial environment meant anything to savers, then Scandinavia appears to have enjoyed a considerable from this point of view. Several factors can explain this contrast but two especially should command our attention. One is structure, the other is policy. As regards the first, Scandinavian economies enjoyed the benefit – which, in southern Europe, Italy had over Portugal and Spain – of a relatively larger not-for-profit financial sector, which was less crisis prone than commercial banks. 22] This was due to several characteristics inherent in such institutions. They had better and cheaper information on the risks posed by clients, lower costs of administration and greater ease in enforcing repayments. Moreover they were less likely to suffer runs by depositors, who knew them well. In addition, they often enjoyed some form of group deposit-insurance, and normally enjoyed deposit guarantees from governments, local authorities or simply groups of local notables.
Lastly, because unlike commercial banks they were not under pressure from shareholders to produce high dividends, they did not have to lend to projects with higher returns but also higher risks. Their image of conservatism more often than not was matched by reality even though they had to contend with the instability that is usually associated with a small scale.  Given how frequent and severe national bouts of financial instability could be, one has to ask whether domestic counter cyclical policies might not have influenced the attractiveness of these financial systems.
As regards government intervention, the low priority given at the time to such policies rules out a significant role for this factor. On the other hand, national banks of issue were just beginning to play the part, informally, of money market regulators, something that would only be enshrined in their charters after the First World War. The timing of their assumption of lender of last resort status matches poorly the way in which these economies responded to financial shocks.
In Denmark and Sweden, two highly stable systems, came to this early, between the 1860s and the 1870s, but so did in Portugal and Italy – the Banca Nazionale, the Bank of Italy’s predecessor was already behaving as a bankers’ bank in the 1860s (Hansen, 1991; Lindgren and Sjogren, 2002; Reis, 1999; Polsi, 1996). On the other hand, Norway and Spain were both latecomers to this field, respectively in the late 1890s and just before the War and yet were at opposite ends of the league table for financial stability (Egge, 1983; Tortella, 1974). 24] While proto-central banking may not have been a major determinant of the closeness between savers and financial institutions, accessibility clearly was. The ease with which economic agents could approach the system mattered a great deal in establishing a relationship with it. One dimension was physical – location, distance, ease of travel – and this was an important reason for the success of the Danish ‘parish savings banks’ (Guinnane and Henriksen, 1998). Another was the suitability for those involved of the institutions available to them.
Savers would more readily supply an institution with funds if, other things being equal, they felt welcome, understood the procedures, knew the people they had to deal with and could easily satisfy burocratic requirements, e. g. minimum size of deposit. In other words, not all institutions and markets served equally well for everyone and this must have had an impact on the propensity to accept the liabilities offered by the system. As Table 8 shows, a considerable disparity existed during these years in the supply of outlets that each system offered. This is one more spect of the enormous distance that separated our two groups of countries, in this case the number of inhabitants per financial outlet, and also brings to light a ranking that matches exactly the ranking presented in table 1. In particular, Italy, which had a density five to eight times that of Portugal and Spain but was five times below the levels of Denmark, Norway or Sweden, had the financial system that managed to attract the greatest volume of savings of southern Europe. Besides the number, the spatial distribution of these outlets also seems relevant.
In Spain and Portugal, financial institutions were an urban phenomenon and were concentrated in the major centres. In Scandinavia, the opposite was the case. A large proportion of financial outlets was in small towns and hamlets. They were therefore close to country people, who were the majority of the country’s population. Thrift institutions were ‘local organizations, formed and run by local people to further what they saw as local goals’ (Guinnane and Henriksen, 1998: 52). . [table 8 about here]
Altogether then, perhaps the most important factor in explaining differences in financial development lay in each society’s propensity to accept non-commercial banking in its midst. Scandinavia’s greater overall capacity to mobilise funds mainly derived from the strength of its thrift sector in all its forms, and this arose because so many people there were prepared to join these movements. Since this cannot be ascribed to major income dissimilarities, nor to diversity in regulation, nor even to the rise of central banking, only two explanations seem to remain available.
Verdier’s (1996) standpoint is that it was political struggles that lay at the heart of the matter. In Denmark, Sweden and Norway, centrifugal political forces were important and prevented the state’s wish to centralize banking, thereby absorbing the resources of the periphery in order to finance central public expenditure. As a result decentralized thrift banking flourished. In the south of Europe, the opposite happened and consequently non-profit banking was suffocated by the pressure from a centralizing state intent on draining the financial resources of the periphery.
There are two objections to this. The first is that what the state wanted these resources for was to finance the public debt, not central public expenditure, and here the contrast between the two regions could not have been greater. As we saw earlier, southern European countries were indeed voracious consumers in this respect. Secondly, we must also not ignore that in these countries finance for the public debt typically does not seem to have come from their peripheries, nor from provincial banks.
Rather, it tended to be held personally and was accumulated at the centre, where saving propensities were apparently higher. This would explain why provincial commercial banking was able to expand in Portugal, Spain and Italy during this period, in spite of their huge public debt commitments (Reis, 2003). Perhaps a more fruitful enquiry should ask why, in the latter countries, except for a small group of wealthy and educated citizens, most of the provincial population appeared remote from institutional saving.
For this we have to try and understand the roots of the stronger impulses in Denmark, Norway and Sweden to create small thrift organisations in terms of the stronger presence of certain social and cultural conditions in these countries (Guinnane, 1994; Galassi, 2000). A basic ingredient was trust, a form of social capital that involved a readiness to accept peer control and to enter into common ventures with other economic agents beyond one’s immediate social circle, which required that direct monitoring and control was left to others.
Societies, like in Scandinavia, founded on a prosperous middle sized peasantry, that experienced successful agrarian reform at the end of the Ancien Regime, and where a more even distribution of income was present, seem to have provided an environment in which such attitudes could flourish. In contrast, apparently this was not the situation in Portugal and Spain, or Italy ‘s south. In Italy, where in some regions only limited liability credit cooperatives (banche popolare) were common, by the 1880s clearly those in the south had much greater difficulty in attracting members’ and their deposits.
As a result, they relied much more on share capital and rediscounts at large banks, and tended therefore to have to hide defaults in their accounts. This in turn exacerbated the problem of trust between members and management, the latter usually from a higher social stratum, and fostered a low-trust equilibrium (A’hearn, 2000). A second element in this approach emerges from the analysis of the early development of modern financial intermediaries, which appeared in Scandinavia already in the first half of the nineteenth century and which very much depended on the degree of human capital endowment present.
Nilsson, Pettersen and Svensson (1999) have shown how literacy in the Swedish countryside before 1850 was associated with the rise in the use of sophisticated credit instruments and probably created a fertile seed-bed for the activity of localised credit institutions. Again, southern Europe was woefully behind in this field, with rates of illiteracy that were still 50 per cent or more in 1900 compared to negligible figures in their northern counterparts. A greater readiness on the part of Danes, Norwegians and Swedes to accept contract money, particularly in small denominations, may be another expression of this cultural dimension. 25] Finally, one should not neglect the influence of the strength of local sentiment in this matter. Its importance has been remarked upon a propos of Denmark (Hansen, 1982), as it has in the case of Italy, the only southern economy where local thrift organizations developed to a significant degree (Polsi, 1996). 5. Why Scandinavian systems recycled liabilities better In comparing national differences in financial intermediation, the second major question of this paper has to do with the efficiency with which the funds gathered by each system were transformed into credits to the private sector.
From this point of view, three circumstances help us understand the efficiency loss of about one half the measures the distance in this respect between the two groups of countries. Possibly the most important one was the more or less chronic difficulty associated with public finance in southern Europe, in contrast to its generally healthy condition in Scandinavia. For the former, this meant a crushing weight of Public Debt holdings and a scant application of these resources to growth inducing purposes. For the second group, not only was this burden far lighter but the resources thus absorbed were also used more effectively for investment.
A further implication, of a more structural nature, was that the problem of an oversized public debt stimulated the emergence in Italy, Spain and Portugal of oversized national banks of issue, which dominated their respective commercial banking sectors, as we saw in section 3, and were the least efficient of all corporate institutions at recycling funds into credits. Why Italy, Spain and Portugal should have been consistently unable to break the grip of budget deficits and of a pyramiding Public Debt is an issue which plunges its roots deeply into the 19th century political, social and military histories of these countries.
Unfortunately, it is impossible to do justice here to such a complex problem. One should note, however, that unstable political institutions, a weak public administration and an excessively powerful military were present in the region throughout the period, unlike what happened in the Scandinavian periphery. As a result, southern European governments found it hard to discipline expenditure, whilst the revenue-to-GDP elasticity tended to be very low as a result of the public administration’s incapacity to increase revenue sufficiently and diversify its sources.
A strong military tended to embroil the country in occasional internal or external costly conflicts that had to be paid for mostly by means of fiscal and monetary unorthodoxy, and was an ever present factor of political instability. Public borrowing in itself was not the difficulty, since the Scandinavians engaged in it without harm. What was dammaging about it to the southern Europeans was its scale, which dwarfed other efforts at mobilising resources, and its use largely to pay off earlier borrowing and maintain a costly ineffective machinery of government.
Negative structural effects on southern Europe were not confined to the politically motivated “excessive” development of their national banks of issue. The socio-cultural reasons that explained the greater prevalence in Scandinavian financial systems of thrift institutions can also be invoked here in explaining their higher global transformation ratios encountered in table 6 above. Less clear, on the other hand, is whether these factors also help explain the fact that, in every type of institution, this region unmistakably led southern Europe in terms of capacity to recycle its liabilities into credit.
Two features of the financial environment appear more helpful in this respect. The first is the difference in the degree of risk that financial institutions had to face. This made it possible for the Scnadinavian ones to immobilise smaller proportions of total assets as reserves, or to avoid tying up resources in safe state bonds. The goals and the quality of management is the second circumstnace to account for differences in the proportion of idle assets in the portfolios of corporate financial institutions.
Possibly, southern European managers were simply reacting reasonably to a riskier investment climate by allocating funds with greater prudence and conservatism than Scandinavian ones needed to. On the other hand, it has been hypothesized (Berthelemy and Varoudakis, 1996: 301) that ‘the technical efficiency of the financial sector is an increasing function of the collected volume of savings [and] that learning-by-doing effects also exist in intermediation activities’.
Poor management has been claimed for both Portugal and Spain (Reis, n. d. ; Sudria, 1994) though comparisons with Scandinavia have yet to be carried out and the case therefore remains open. As regards Portugal in particular, it has been shown, following Hinderlitter and Rockoff (19.. ), that, after taking risk differences into account, the share of unused funds in the balance sheet of commercial banks was greater than could be justified by reference to practices in contemporary major financial centres.
Finally, the high returns on government issued liabilities in southern European caused resources to be diverted away from private credit operations in contrast to Scandinavia where the yield of such holdings was comparatively less attractive and better alternative investment opportunities seem to have been more numerous. 6. Conclusion During the course of the ‘long second half of the 19th century’, the southern and the northern peripheries of Europe followed contrasting paths of financial development.
This led to quite disparate results in the supply of credit to the non-state non-bank part of their economies and justifies perhaps speaking of a “Scandinavian”, as opposed to a “southern European” type of financial system. Having quantified these differences, this paper argues that the gap is large enough to justify the view that finance contributed to the divergence in economic growth between the two regions. To address the reasons for the substantial efficiency differential between southern European and Scandinavian financial systems, it was necessary to break this down into the two basic functions that financial systems carry out.
One concerned the mobilization of savings as financial liabilities of these systems. The second revolved around the conversion of these liabilities into credits to the non-financial private sector. Southern European countries were losers in both instances. The analysis of the first of these brought to light that Scandinavian institutions were capable of mobilizing comparatively more resources than their southern counterparts, the exception being in state bonds, where the latter led by a clear margin.
The second dimension of this study revealed that Scandinavian institutions were also capable, type by type but equally in toto, of extracting a larger quantity of credit from their laibilities in order to make them available to the productive sectors of the economy. To some extent these contrasts were caused by endogenous conditions. The simple fact is that Italy, Portugal and Spain were consistently poorer and were becoming increasingly so. This affected their volume of savings but also probably lessened the demand for the financial outlets that enabled savers to recycle these funds as institutional financial liabilities.
On the other hand, exogenous factors probably also played an important part in helping to understand fully the process of financial development in these two regions. Essentially, three aspects are involved here. Regulatory conditions have loomed large in many analyses of this type but do not appear to have had a significant impact on the global outcomes picked up here, although they probably shaped some of the structural differences observed. A much stronger case can be made instead for the part of political, social and cultural factors in driving a wedge between the financial development paths that we have observed above.
The last two were instrumental in leading to a greater development in Scandinavia of the non-commercial bank sector. This was responsible for diversifying the supply of financial outlets, attracting an much greater volume of savings per capita and then ensuring that a larger proportion of such funds became available for investment purposes. In all their complexity, political factors probably mattered most of all because they translated savings into a huge mass of state issued liabilities that stifled the expansion of the other parts of the system in the southern countries.
At the same time, having helped to mobilize these funds, politics then became responsible for their sterilisation as financial instruments. When all is taken into account, it is this which perhaps explains the best part of the great financial divide between our two sets of countries. Of politics it can always be said that it might have been otherwise and this may seem a trivial conclusion. To claim this, however, would be to ignore the fact politics and institutions have long histories too and that path dependency is not solely the preserve of economic phenomena. References A’Hearn, B. (2000).
Could southern Italians cooperate? Banche Popolari in the Mezzogiorno. Journal of Economic History 60, pp. 67-93. Aleotti, A. (1990). Borsa e Industria. 1869-1989: Cento Anni di Rapporti Difficili. ??? : Edizioni di Comunita. Beck, T. , Levine, R. and Loayza, N. (2000). Finance and the sources of growth. Journal of Financial Economics 58, pp. 261-300. Berthelemy, J. C. and Varoudakis, A. (1996). Economic growth, convergence clubs and the role of financial development. Oxford Economic Papers 48, pp. 300-28. Broder, A. (1976). Les investissements etrangers en Espagne au XIXe siecle: methodologie et quantification.
Revue d’Histoire Economique et Sociale 54, pp. 29-63. Bruck, C. (ed. ). (1995). Les Caisses d’Epargne en Europe. Paris: Editions de l’Epargne, 2 vols. Calomiris, C. (1995). The Costs of Rejecting Universal Banking: American Finance in the German Mirror, 1870-1914 in N. Lamoureaux and D. M. Graff (eds), The Coordination of Activity within and between Firms (Chicago: University of Chicago Press). Cameron, R. (1967). Banking in the Early Stages of Industrialization. New York: Oxford University Press. Cameron, R. (ed. ). (1992). Fina