The only way to grow? Italian Business groups in historical perspective

ABSTRACT This article analyses the dynamics of business groups (BG) formation and diffusion in Italy in the twentieth century. It shows that BGs is not an organisational form which characterizes only developing countries or economies in their very early stage of development. Indeed, in its evolution from a peripheral country to one of the most advanced economies, Italy has been constantly populated by BGs. One striking feature of the Italian corporate system is that BGs are present not only among large firms, but also in almost all the other forms of enterprise: cooperative firms, municipalised companies, state-owned enterprises (SOEs), and fast-growing medium-sized firms. Thus, BG seems to be the ‘only way to grow’ for Italian firms or, at least, the easiest way to reach a reasonable size. BG as a governance system looks particularly flexible, adapting itself to different ownership, market conditions and local contexts. In absence of obstacles of legal or fiscal nature, this flexibility is probably the main reason for its resilience.


Introduction
An article in the Financial Times ('Italian Business: No Way Back, ' 21 August 2013) suggested, in quite an optimistic way, that the traditional, collusive attitude of Italian big business would be progressively crumbling under the pressure of international institutional investors. In a colourful way, the columnist used the image of the salotto buono (the fine-drawing room) to indicate the clubby, exclusive atmosphere surrounding the top-layer of Italian capitalism. Since the start of the twentieth century, indeed, as stressed by most of the available research, the prominent Italian companies tended to be interwoven with a dense network of cross-shareholdings, creating a coordinated structure allowing top owners/managers to keep firm control over their companies with limited investments. While the cited article focuses mainly on the role played by Mediobanca (until a few years ago Italy's sole investment bank) and Assicurazioni Generali (Italy's largest insurance company) in designing and maintaining this architecture, research about Italian capitalism has stressed how, alongside this cluster of companies (resembling a sort of 'horizontal' keiretsu), many vertical concerns existed.

ABSTRACT
This article analyses the dynamics of business groups (BG) formation and diffusion in Italy in the twentieth century. It shows that BGs is not an organisational form which characterizes only developing countries or economies in their very early stage of development. Indeed, in its evolution from a peripheral country to one of the most advanced economies, Italy has been constantly populated by BGs. One striking feature of the Italian corporate system is that BGs are present not only among large firms, but also in almost all the other forms of enterprise: cooperative firms, municipalised companies, state-owned enterprises (SOEs), and fast-growing medium-sized firms. Thus, BG seems to be the 'only way to grow' for Italian firms or, at least, the easiest way to reach a reasonable size. BG as a governance system looks particularly flexible, adapting itself to different ownership, market conditions and local contexts. In absence of obstacles of legal or fiscal nature, this flexibility is probably the main reason for its resilience.
Since the beginning of the Second Industrial Revolution, with the rise of the technological trajectories linked to electricity and chemicals, large companies tended to organise themselves as hierarchical business groups (BGs). Although the definition of BGs is somewhat debated, we define it as the peculiar corporate form based upon a 'parent' company holding control stakes in and exerting a co-ordination activity over subsidiaries. 1 In Italy, these groups were normally headed by holding companies, often through sophisticated pyramidal structures able to amplify the benefits of financial leverage. This collusive attitude was thus permeating almost all the country's large companies, both private and State-owned, which in this respect did not behave differently from the purely private ones. The rationale behind this strategy was to maximise control (and often tunnelling of resources) limiting the amount of resources directly invested by the owners/controllers, an absolute necessity where big business in capital intensive industries was concerned. To describe this, one of the most acute observers of Italy's industrial history, Napoleone Colajanni, coined the effective locution 'capitalism without capital' . 2 Italian business historians have so far analysed an increasing amount of evidence about this behaviour, pointing out the (few) pros and (many) cons of such a structure, 3 suggesting how this collusive attitude has been somehow endogenous to Italian capitalism, as a result of a dramatic institutional failure in determining clear and transparent 'rules of the game' . 4 However, this criticism has almost never been applied to forms of enterprise other from large private and stateowned firms. A large section (the majority, in numerical terms) of Italian companies has been thus genuinely immune from the growth strategies characterising big business. For instance, network-like structures have been considered as distinctive characteristics of small firms in industrial districts, but as a (positive) symbol of the co-operative attitudes among local producers, bounded among themselves by both social and economic ties. 5 At a closer look, however, this conventional wisdom shows some deficiencies and a different picture emerges. As this article shows, vertical BGs can today be found almost everywhere in the Italian industrial landscape, and involve companies of different size, legal status and sectors. Instead of vanishing or weakening, the BG, as a governance and organisational device, seems to further root and diffuse among Italian enterprises as time goes by, that is over the nation's whole industrial history. All this calls for a deep investigation of the reasons which lie behind this stubborn persistence and growing relevance. This article aims at providing an in-depth analysis of the typologies, dynamics and rationales for the growing diffusion of BGs in Italy. In doing this, we combine the longitudinal approach typical of business history with a qualitative analysis providing a tentative taxonomy of the various typologies of BGs. Combining this evidence with that emerging from previous research, 6 we put also forward some estimates of BGs' quantitative relevance over time.
This article is structured as follows. After this introduction, Section 2 presents an outline of the evolution of the Italian corporate system since the outset of Italy's modern industrialisation in late nineteenth century. This provides a crucial framework for Section 3, which places an articulated taxonomy of BGs in the various forms of enterprise that can be found across Italy's industrial history. Section 4 explores the rationales for the wide use of BG, both across time and in each form of enterprise. Section 5 concludes.

The evolution of the Italian corporate system
Even if a latecomer in modern economic growth, Italy followed the global trends in the evolution of contemporary capitalism across the technological paradigms of the three industrial revolutions.
The creation of German-type universal banks in the 1890s enabled Italy to catch up in capital and energy-intensive industries. However, in the early 1930s, Italian big business was tremendously hit by the great depression, and the largest universal banks found their balance sheets heavily burdened by toxic assets: credits towards the distressed industrial systems, plus the shares of failing big business. The solution, a giant bailout of the nation's three largest banks, brought in the hand of a state agency, the Istituto per la Ricostruzione Industriale (IRI), a considerable share of the whole share capital of Italian stock companies. In 1936, a new Banking law allowed banks to provide only short term credit to the industrial sector, explicitly prohibiting them to act as shareholders and to provide long and medium term credit to industrial firms, which was left to State-owned agencies set up with this explicit purpose. 7 At the same time, the stock exchange, was seriously weak and not attractive for both companies and investors. 8 On the one hand, the Italian stock exchange was structurally lacking of any form of regulation and protection of minority shareholders. On the other hand, the weak regulatory framework allowed the main blockholders (the state, individuals and families) to leverage on control-enhancing mechanisms and low transparency in order to keep high control levels with limited investment. This turned out to be a formidable incentive to firms to grow by creating BGs rather than widely-held M-form corporations. 9 In the three decades following the Second World War, such a huge complex of Stateowned enterprises (SOEs) and a state-controlled banking system speeded up the growth of the industries of the Second Industrial Revolution. 10 As a result, the weight (as a share of total output) and industry composition of Italian big business converged with those of the most industrialised countries. 11 The oil crisis of the 1970s and the advent of Information and Communications Technology (ICT) lowered the process of convergence of Italy's industrial structure with that of the leading nations. Subsequently, divergence started. 12 The weight of big business -both in term of total assets and employment -decreased in Italy more than in the other advanced economies. 13 Divergence concerns also the distribution of large firms across sectors, with a much smaller weight of ICT-related industries in Italy than in other developed countries, and, more generally, a lower capacity of firms to innovate. 14 The decline of big business since the 1970s was accompanied by the rise of other forms of enterprise: small firms (often clustered in industrial districts), co-operative firms, municipalised firms, medium-sized enterprises (the Italian Mittelstand), and foreign-owned firms, which increased their share of both total employees and manufacturing output. 15 Some of these forms of enterprise are strictly specific to the Italian case: small firms (from which most of the Mittelstand came out), co-operatives and municipalised companies correspond, in their diffusion, form and structure, to a particular set of social and institutional conditions which characterised twentieth century Italy. Conversely, other forms, such as big business and foreign-controlled firms, are common to the other industrialised nations. However, Italian big business remained sharply different from its American and European counterparts as regards the degree of internationalisation, the ownership and organisational structures, and the governance practices. 16 By contrast, a wide range of policies was put in place to foster the expansion of the small firm sector, co-operative firms and municipalised companies. 17 As previously mentioned, the pervasive presence of the BG is a common trait of growth of all forms of enterprise in Italy. In the next Section we show how the rationale for the use of the BG form constantly changed across forms of enterprise and over time according to technological and institutional changes.

Typologies of business groups
The main aim of this section is to offer a broad overview of the large diffusion of BGs in the Italian economy, providing a "new" taxonomy of groups' typologies. In Italy, the pervasive presence of BGs is not limited to big business, as usually happened in other countries, but it is common to most of the other forms of enterprise. Considering a set of variables, such as size, performance, legal forms, types of ownership and governance, in a recently edited book, Colli and Vasta identified seven different forms of enterprise which have contributed to the competitiveness of Italian capitalism along time. 18 In all of these, but the foreign-owned, the BG structure is largely dominant and their weight on the total of the Italian joint stock companies is rather impressive. Colli and Vasta provide a rough estimation of the quantitative relevance of big BGs in the Italian economy. According to their calculations (see Figure  1), the first 15 BGs (both private and state-owned) 'controlled' on average two thirds of the domestic joint stock companies' total assets and more than half of their share capital. 19 looking at Figure 2, which shows in the x-axis the ownership characteristics of the firms and in the y-axis their size, we can see all the forms of enterprise taken into account in this section. In the upper part of the figure, we find the forms of enterprise for which the BG presence has been largely investigated. Firstly, we find the private big business (I quadrant) for which BGs have been largely dominant since the beginning of the twentieth century (section 3.1). In the II quadrant there are SOEs, which have represented, from the 1930s to the 1990s, a noticeable characteristic of Italian capitalism (section 3.2).
In the lower part of the figure, we find the forms of enterprise for which the presence of BGs has been underestimated by the literature. Close to the centre of the figure, we find the co-operative firms which -originally characterised by small size and considered as an alternative to the 'capitalist corporation' -since the 1980s have developed peculiar BG structures (Section 3.3). In the III quadrant, there are the municipalised firms which, despite having preserved a public ownership, showed since the mid-1990s a growing propensity to be organised as BGs (Section 3.4). Finally, in the IV quadrant we placed the small firms, from whose seedbed arose, the Mittelstand after the 1980s, characterised by a large diffusion of BGs (Section 3.5).

Private big business
As anticipated in the previous section, even though a latecomer country, Italy closely followed the same pattern of development across the first and second industrial revolutions which characterised the other industrialised countries. until the 1970s, alongside a vast population of small firms, big business active in capital intensive industries took root and accounted for a growing share of the nation's industrial output. Since the 1880s both domestic and foreign entrepreneurs started successful ventures in fast-growing industries such as steel (Falck, founded in 1906), caoutchouc (Pirelli, 1872), electricity (Edison, 1889) and motor vehicles (Fiat, 1899). The process of growth was, since the beginning, characterised by a strong tendency towards the progressive agglomeration of companies, more than pure processes of 'internal growth' .
Thus, agglomerations of companies resembling BGs were formed since the beginning of the twentieth century, mainly in the form of cartels. 20 A dense web of cross-shareholdings and interlocking directorates was designed to strengthen the mutual agreements among the companies belonging to the cartel, as for instance in the case of the 'steel trust' , created in 1903 by the main iron and steel producers which were in turn coordinated by Italy's two largest German-style mixed banks, the Banca Commerciale Italiana and the Credito Italiano, with the aim to control a relevant portion of the nation's steel production. 21 As in other cases, 22 trusts and cartels, in this phase, had as their main rationale the control of competition although this might lead to the survival of inefficient companies. The result was the diffusion of 'horizontal' cartels of collusive nature, often coordinated by the two largest mixed banks, not only in steel, but also in cotton, 23 sugar refining, 24 and even in beverages production. 25 Military procurement for the First World War was for many of the existing 'first movers' a unique opportunity for expanding the scale and scope of their activity. As a result, the leading industries of the second industrial revolution acquired a clear oligopolistic structure. 26 The process of concentration was, almost without exceptions, managed through the creation of a standard-type vertical BGs, i.e. as previously mentioned, a collection of legally independent entities linked among themselves and to a 'parent' company -usually controlled by a family or by a coalition hinging on one of the largest mixed banks -by cross-shareholdings and interlocking directorates. Examples range from Montecatini, a mining company, which Figure 3 the process of expansion of edison, the italian leader in electricity production and distribution (1912)(1913)(1914)(1915)(1916)(1917)(1918)(1919)(1920)(1921)(1922)(1923)(1924)(1925)(1926)(1927). source: adapted from Pavese, "Le origini, " 142-143. the size of circles represents the size of nominal share capital of single companies in the group.
undertook an ambitious downward integration, becoming the leader in fertilisers, to Fiat which became a powerfully integrated engineering and motor vehicle company, and to Ansaldo, a huge vertically integrated 'konzern' in steel and heavy mechanics. 27 In this phase, vertical BGs of hierarchical nature started to diffuse together with horizontal ones. Moreover, the top management of the country's largest banks put in place sophisticated entrenchment strategies. The banks thus increased their shareholdings in a number of industrial companies, which, in their turn, were encouraged to buy shares of the banks. As a result, other horizontal, collusive groups emerged, pivoted by the main banks. 28 In the interwar years, Italy made consistent progress in capital and energy-intensive industries of the second industrial revolution. In the 1920s, the vast majority of BGs continued to be formed through the progressive aggregation of companies, listed and not listed, active largely in the same industry or in related fields, very often in a logic of technologically driven vertical integration. Several examples can be found in the electricity industry. Here the concentration process took place through the acquisition of smaller companies (frequently active on a local or regional basis) by larger ones, which resulted in an oligopolistic structure on a super-regional basis. An exemplary case is that of Edison ( Figure 3).
Even industrial companies expanded in the same way, both during and after the First World War. One (even if not the sole) driver of this process was the decision of de-taxing war super-profits of companies when reinvested, which strongly encouraged entrepreneurs to pursue 'external growth' strategies. A case in point here is that of Fiat which expanded both through internal (a new, ultra-modern factory, the lingotto, was inaugurated in 1922) and external growth through the acquisition of controlling stakes in iron and steel companies in Piedmont already belonging to a cartel, the Gruppo Piemontese. In chemicals, immediately after the First World War, Montecatini started a steady expansion through the acquisition of its main competitors, Società Prodotti Chimici Colla e Concimi and unione Concimi. During the following decade, Montecatini pursued an intense policy of acquisitions, both in related and non-related activities, due also to the pressures of the Fascist regime which encouraged the chemical concern to rescue companies on the verge of bankruptcy in exchange of protection in the internal market for fertilisers. 29 After the Second World War, the process of growth of big private companies went on along these lines. During the late 1950s and the 1960s private big business expanded basically through a mix of investments and endless acquisitions, which enlarged the boundaries of some of these groups enormously. Thus, the major electricity producer, Edison, was by the beginning of the 1960s (when the whole electricity industry was nationalised) a corporate 'Moloch' composed of more than one hundred companies and subsidiaries, by far Italy's largest group in terms of revenues and market capitalisation. 30 During the 1970s, a period characterised by crisis and the re-structuring of large companies, new BGs emerged, and others reshuffled their corporate architectures, for reasons completely different from those which drove BG expansion during the 'Golden Age' . large companies in capital intensive industries were heavily affected by the crisis in their cost structure. In order to gain more flexibility, minimise control costs, dilute tensions in labour relation (and in some cases to proceed more easily with plant closures) most vertically-integrated companies followed a strategy of decentralisation and fragmentation of the production chain. This practice was also encouraged by public incentives, which allowed each single subsidiary of a same group to have access to public funds for investments, as happened in the case of Società Italiana Resine (SIR), a chemical company which spun off hundreds of subsidiaries in order to obtain funds from the State for its investments in Sardinia. 31 During the 1980s, big private firms continued to heavily rely on the BG form as common denominator in their corporate architectures. Again, the determinants of this process changed. First, the system of corporate taxation was modified, and intra-group transactions were no longer taxed -something which reduced incentives to corporate integration. The second determinant was mainly of an exogenous nature, and involved a mounting effervescence in the stock market. Spinning off and listing subsidiaries thus became the most viable strategy, followed by all the major private groups Fiat, De Benedetti, and Ferruzzi, in order to gather additional resources from the market. The configuration of Italian BGs was heavily influenced by this process. Pyramidal structures now became generalised; pure financial holdings, in their turn directly controlled by individuals or families, controlled a chain of operating companies, some of which listed. 32 This process of unrelated diversification was in some way an 'Italian' version of the conglomeration wave ongoing at international level. In Italy, this process went on longer until the 1990s. Starting in this decade the weight of private big business in the Italian corporate system fell dramatically. Some BGs that played a leading role in the nation's industrialisation -such as Olivetti, Falck and Montedisonfailed and were dismantled. 33 The remaining BGs further enlarged their boundaries thanks in particular to a massive sell-off of SOEs. In some cases, privatisations led existing groups to pursue unrelated diversification aimed at re-orienting their business towards the domestic market. This was the case of Benetton that acquired Società Autostrade, Italy's largest motorway company. 34 However, other groups made their acquisitions in related activities, i.e. in steel (lucchini, Riva, and Marcegaglia) and in chemicals (Mapei). 35

State-owned enterprises
The capital and scale-intensive sectors (steel, electricity and chemicals) typical of the second industrial revolution started their consolidation in the decades following the First World War. The role of the state in this process was very important, as the state, since the unification, supported industrialisation through military procurement, protection and tariffs. The state's direct involvement in the ownership of companies started in 1933 with the creation of the IRI, originally conceived as an emergency measure. However, after only few years, in 1937, IRI became a permanent institution and represented, for nearly 70 years, one of the pillars of the Italian corporate system. IRI was created in order to bail-out Italy's three largest mixed banks (Banca Commerciale Italiana, Credito Italiano and Banco di Roma), which were on the verge of collapsing due to the great crisis of the early 1930s. IRI took over the banks' shareholdings and became the majority owner of a multitude of companies in a vast array of industries: i.e. real estate, steel, shipbuilding, mechanics, textiles, and utilities. The weight of this group in the Italian economy soon became enormous. Already in 1936, when it was still temporary, the IRI group accounted for 12.5% of the total capital of Italian joint stock companies. 36 A complex architecture was designed to give some order to the vast array of companies under the control of IRI. The group, as shown on Figure 4, which represents an example of its structure without referring to a particular period, was based upon a 'super-holding' (IRI) under the direct control of the government. The super-holding controlled some financial holdings which were characterised by sectorial specialisation: i.e. Finsider (steel), Finmeccanica (mechanics and engineering), Finelettrica (electric energy).
The boundaries of SOEs further expanded after the Second World War: IRI remained the main pillar of the system but a second pillar, the state energy super-holding, Ente Nazionale Idrocarburi (ENI), was founded in 1953, prolonged the experience of AGIP already founded in the 1920s. 37 The weight of IRI group rose considerably and, in 1952, it represented almost one fifth (19.4%) of the total capital of Italian joint stock companies, while ENI group accounted for 3%. 38 These two big state-owned groups contributed considerably to Italy's catching up to the technological paradigm of the mass production during the 'Golden Age' . IRI gave a considerable contribution in the creation of infrastructures and heavily invested in capital-intensive industries, especially steel. Moreover, IRI also tried to provide inter-dependencies across companies in different industries. 39 At the same time, ENI, which had a similar vertical organisation, reinforced by interlocking ownership across the companies of the group, was directed to meet the country's needs in the fields of energy. State-controlled groups showed a number of differences from private ones. These concerned in particular: (1) the use of more sophisticated multi-level organisational architectures; and (2) the presence, particularly inside IRI, of strong linkages with the country's largest banks. 40 In the 1970s, the oil crises and the rise of a new technological paradigm, based on ICT, opened a new scenario for the Western economies. Italy was severely hit by this new situation and SOE groups were increasingly burdened with special social objectives, such as prompting the industrialisation of the backward South or rescuing distressed companies. 41 The structure of Italian corporate system turned noticeably after the oil crisis and the new technological paradigm speeded up the crisis of the SOEs. This degenerated in the 1980s due to the excessive grip of politicians on SOEs and the consequent erosion of the 'mission' culture of SOE managers. 42 Somewhat paradoxically, at the beginning of the 1980s, the boundaries of the SOE sector were the largest in Italian history amounting to about 40% of the total share capital of Italian joint stock companies, 43 but the final crisis of the entire system was close.
During the 1990s, massive privatisations reduced the weight of SOE groups opening a new era of state entrepreneurship in Italy. IRI was dismantled and closed down even if the state retained control of national champions in some strategic sectors as energy, telecommunications and defence. The stock of these companies was also partially sold off to raise funds from private investors. At the end of the century, the diffusion of state owned groups within the Italian corporate economy was considerably reduced.

Co-operatives
If the diffusion of BGs among Italian big business, both private and state-owned, has been widely investigated, this was not the case for BGs in other forms of enterprise. Among the latter, co-operatives stand out. The first co-operative firms in Italy were founded in the late nineteenth century. The plurality of inspirational ideals in the Italian co-operative movement led to the formation of separate umbrella organisations: legacoop (leftist), Confcooperative (catholic), and Agci (liberal). 44 Their tasks are manifold: new co-operatives promotion, strategic marketing, ethical control of the affiliated firms, know-how and human resources circulation, internal conflict resolution and coordination of the major strategic decisions, lobby at the local, national and European government level. 45 until the 1970s, co-operatives played a marginal role in the Italian economy: nearly all co-ops were small and inefficient and survived only thanks to political protection. However, since the 1970s the co-operative sector -contrary to the other major advanced economies 46 − went through constant expansion: from 10,000 co-ops in 1971 to 63,000 in 2010; from 2% of the nation's employment in 1971 to 6.4% in 2010, a value higher than in any other European country. 47 The larger co-ops experienced the most outstanding growth. As a result, in 2010 co-ops accounted for 10% of all the enterprises with more than 500 employees. larger co-ops have become the market leaders in Italy in some sectors, i.e. large scale retailing, construction, and agro-food. Such a result was probably favoured by the capitalist sector's (especially private big business') failure to achieve the power retained in other countries, which left co-operatives with more room for expansion. 48 The growth of larger co-ops was first pursued through mergers of existing co-ops and then, since the mid-1980s, above all through the formation of BGs that encompassed both co-operative firms and joint stock companies controlled by a co-op. This growth strategy was adopted principally to overcome capital constraints. In fact, under-capitalisation was the major weakness of co-operative firms, which constrained their growth. This was due to the lack of a market for co-operative shares, from which the expected future flows of profits could increase their value for the shareholders. A paramount role to boost the creation of co-operative groups was played by the 1983 law that granted co-operatives the permission to own capitalist corporations. This allowed many co-ops to grow by becoming the parent-company of several joint-stock companies which were used to attract capital from the market. 49 Thus, in 1998 there were 64 co-ops as members of legacoop (Italy's largest umbrella organisation) that presented a consolidated balance sheet; in 2003 the corresponding number rose to 96. The value of their shareholdings surged from 1354 to 2793 million euros. 50 An exemplary case is Granarolo, a consortium set up in the 1950s by some milk producing co-ops of Bologna. Its plants, machinery and brands were owned partly by the consortium and partly by the milk producing co-ops that had constituted it. In 1991, they created a joint-stock company with the same name, to which the assets of the consortium and of the co-operatives were transferred. This became the company designing the strategy for the entire group. In 1998, the original co-operatives were gathered in a single consortium, Granlatte, which became the owner of the joint stock company, together with a second consortium, Cooperlat, which joined in 1999. In this way, it was possible to resort to more substantial bank loans and to issue bonds, without the traditional disadvantages produced by the co-operative legal form. By 2010, the group was articulated on four levels, of which the second was made up of two cooperative consortia, which, directly or indirectly, owned 11 companies. The group maintained its cooperative character, because 68% of the milk processed came from its member co-operatives. In 2004, Banca Intesa, one of Italy's largest banks, acquired 19.8% of the capital of Granarolo. 51 Most co-operative BGs pursued related diversification and these were the most successful. There were also some cases of unsatisfactory unrelated diversifications, as Cmb of Carpi, a co-operative construction giant which in the 1980s diversified in the ailing local textile industry. This venture ended in a massive loss as Cmb lacked the know-how to manage the newly acquired textile subsidiaries. 52

Municipalised companies
Another form of enterprise typical of the Italian corporate system appeared at the beginning of twentieth century. The 1903 Municipalisation Act enabled municipalities to manage public services through the Azienda Municipalizzata (municipalised company), an autonomous public-law entity emanating directly from the municipality itself. Since then, many municipalities, especially in the North and the Centre, set up municipal companies to provide services such as electricity, gas and water supply or public transport acting as monopolists in the territory of the municipality.
The post-Second World War years saw a substantial expansion of the area of municipalised companies, prompted principally by the local governments controlled by left-wing parties. Nonetheless, municipal companies remained small as their growth was curbed by the impossibility of operating outside of the territories of the municipalities that owned them. 53 In 1990, a new law on local Authorities spurred many important changes in the municipal companies. The law considered the Azienda Speciale (special company) -whose managers enjoyed greater control of the firm -as the most suitable legal and administrative form for managing local public services. It was directed towards the rationalisation of management strategies and instruments, thus indicating a range of privatisation solutions. 54 In the 1990s, new legislation was also passed in the utility sector to introduce competition in the retail segment and a new regulation of access to market segments requiring significant investments through competitive tendering procedures. Municipalised companies could now sign contracts to obtain energy supplies at lower prices and could consider managing other city and regional distribution networks. Partnerships with other private or public companies allowed them to enlarge their range of action into other territories from the original ones. At the same time, customers were no longer guaranteed and had to be continuously won, especially industrial ones. 55 This new competitive environment pushed many municipalised companies to turn into public limited companies that pursued multi-utility strategy aimed at diversifying services and enlarge their territorial base. Municipalities remained the most important block-holder in these companies, followed by private owners and by regional, provincial and state organisations. Some of the new multi-utilities were listed in the Stock Exchange, starting with Amga of Genoa in 1996. A process of consolidation of municipalised companies led to the formation of BGs that operate at a regional or supra-regional level. These are constituted by a parent-company (usually a holding company) which controls several operating companies. The latter specialise in the managing of the individual utilities, in upstream or downstream activities, or in serving the territory of the single municipalised companies that joined the BG. 56 The BG form not only proved very effective in increasing the leverage to finance related multi-utility diversification and territorial expansion, but also in accommodating the pressures of the shareholders and stakeholders of the territories (the municipalities) in which the constituent companies (the municipalised companies) were located. This has not been devoid of repercussions on the performance of such companies. In fact, a recent study has shown that the large presence of politicians in the board of directors of Italian utilities has an inflationary effect on employment and harms the accounting performance of these companies. 57

The Italian Mittelstand
Starting from the late 1980s, an important evolution in the Italian economy has been the growing relevance of medium-sized enterprises, the Italian Mittelstand. At the beginning of the twenty-first century, more than 1500 businesses with annual sales between 150 million and 1.5 billion euros operated in Italy. These Mittelstand companies often emerged from the entrepreneurial seedbed of industrial districts. In fact, the increasing pressure of globalisation prompted a restructuring of many Italian industrial districts, with the rise of lead firms that pursued dimensional growth and accelerated internationalisation with the establishment of trading and production subsidiaries abroad. Most of these Mittelstand companies became leaders in world-sized market niches, especially in traditional sectors such clothing, luxury goods, and various branches of the so-called 'Made in Italy' sector. In the global competition, these companies are nowadays mainly characterised by the low wages of their employees and a scarce attitude to develop formal innovation processes. However, it is worth saying that in several cases there are also firms specialising in machine tools or in other more advanced sectors.
Generally, these companies went on sharing many of the features of the traditional small firms typical of industrial districts, i.e. family ownership, centralised management and a scarce attitude towards path-breaking innovation. 58 The emergence of this model as a successful form of enterprise is someway puzzling because, in a long-term perspective, 'medium-ness' is a transient condition. By looking at the whole twentieth century, only a very few Italian firms found their optimal size within the medium-size group: it seems to confirm the traditional difficulty to grow faced by many Italian firms which, after having become medium-sized, in most cases returned to being small. 59 The growth of Mittelstand companies in recent years was pursued through the creation of BGs based upon a multi-subsidiary system. The owner-family is the source of skills and resources most of these BGs have relied upon to provide finance, labour, and know-how. The structures of these BGs mirror their historical evolution. The first, original nucleus of the enterprise has transformed itself into a family-controlled financial holding company, presiding over a large number of unincorporated subsidiaries active in related activities, both in the home market and abroad. The BG form is the result of both the minimisation of administrative and coordination costs and of the growth strategy pursued by these enterprises. In fact, most Italian Mittelstand companies grew (both in the domestic and foreign markets) usually by means of the acquisition of existing smaller companies which maintained their strategic and administrative independence. The operative companies (both productive and commercial) are often run by relatives or by professional managers strictly linked to the family to whom is often assigned also a minority shareholding. In other cases the external growth is pursued leaving the existing management (sometimes the former owners) with the responsibility of running the activity. 60 Thus, the rise of these Mittelstand BGs appears to be the result of the internationalisation strategies adopted by the most dynamic firms to face globalisation. This led to an original blending of technologies of the first and of the third industrial revolution. In fact, most of such BGs have their core business in the typical first industrial revolution 'Made in Italy' sectors, but in order to pursue successful internationalisation they massively used ICT-based technology to adapt to rapidly shifting markets and to manage offshore outsourcing as well as a worldwide network of trading subsidiaries. 61 In this case, the rationale for the use of the BG form was not the increase of leverage, as is shown by the fact that the architectures of such BG are usually very simple, most of these BGs have only a two-layer structure (a holding companies directly controlling several operating companies). 62 Instead, the creation of BG served principally to acquire, retain, and properly use technical and managerial capabilities, by giving the managers of the subsidiaries a wider autonomy than the division heads of a M-form enterprise.
An exemplary case is that of Wam, a firm set up in 1968 as an owner-run company by Vainer Marchesini -at that time a 22-year-old technician -in the mechanical engineering district of Modena to produce cement screw conveyors. Starting from the 1980s, Wam pursued an internationalisation strategy that enabled it to become the world leader in the production of bulk material handling and dust filtration machinery. By 2007, the company was still run by its founder. The holding company Wamgroup -owned by the Marchesini family -was at the head of a BG which consisted of 45 companies: seven Italian manufacturing companies, nine foreign manufacturing companies (in the uS, China, Australia, India, Brazil, Spain, Romania and Croatia), 27 foreign trading companies, one Italian real estate company and one Italian service companies. The group employed 1700 workers and had a turnover of about 200 million euros. 63

Why growth through business groups?
After having proposed a taxonomy of different BGs, this section discusses the reasons why Italian entrepreneurs (in all forms of enterprise) preferred to grow by creating BGs rather than trying to develop large M-form corporations.
A first reason refers to BG as an instrument to enhance the control of the dominant shareholders with a limited investment of their own resources, an explanation which goes back to Hilferding. 64 The diffusion of such a control enhancing mechanism implies in its turn the presence of a favourable legislation, or at least the absence of rules limiting the leverage effect. Above all -and in the case of Italy this seems particularly evident -the limits in the dimension and in the quality of the domestic financial markets, and the absence of specialised financial institutions, play a relevant role in incentivising the formation of BGs through which the dominant shareholders obtain resources from the stock market without losing much of their decisional power. This was the main reason that pushed private big business and SOEs to create multi-layered BGs to catch up with the technological paradigm of the second industrial revolution and, more recently, for the creation of BGs by co-operatives and municipalised companies.
Secondly, BGs provide more freedom and incentives to the subsidiaries' management, something which is more difficult in an integrated organisation in which the headquarters has a strong grip on the divisions and in which the management's decisional power is severely constrained. This becomes particularly evident in the case of the Italian Mittelstand, due to the key role played in it by the owner-family: BGs allow separate power and responsibilities across the members of the family, in this way avoiding conflicts among them. In other cases, the BG is an instrument through which is possible to take over companies without integrating them, and avoiding the risks of a complete turnaround in their managementespecially in the case of successful ones.
A third possible explanation for the diffusion of BGs lies in the fact that this form proved very effective in accommodating the pressure of both local shareholders and stakeholders. This was particularly the case for BGs that were formed in the public utilities from the aggregation of the former municipalised companies. In this case the setting-up of a separate subsidiary -with its own staff and board of directors -in each local context (a province or a union of municipalities) it is used by the BG to create new offices for individuals connected with local politics and, as a wider scope, for the interaction between the BG and the local community.
In other cases, fiscal policies play a relevant role, especially when they neutralise the difference between BG and other forms of organisation (i.e. the M-form). large BGs can have a better access to credit than large integrated organisations, at least from the profitable subsidiaries. Moreover, the BG structure better allows the tunnelling of resources from companies where the holding has low cash rights to companies where it has high cash rights. Another peculiar situation is where the multiplication of subsidiaries is due to the possibility of obtaining resources in the form of incentives and subsidies from public institutions. In the case of Italy, the policies aiming at fostering the industrialisation of the South were largely based on the concession of tax exemptions, loans, and also of direct subsidies. This led some entrepreneurs to create BGs characterised by a wide number of subsidiaries, each one separately asking for public support.
In sum, the BG as a governance system looks particularly flexible, adapting itself to different ownership and market conditions. In absence of obstacles of legal or fiscal nature, this flexibility is probably the main reason for its resilience. This points to the institutional determinants of its adoption, i.e. why the policymaker incentivised, or at least did not hinder, the formation of BGs. An analysis of this is beyond the scope of the present article. We can suggest that this was probably a consequence of the fact that mass-based political parties soon became the most prominent political actors in Italy, which profoundly conditioned institution building. These parties prevented the creation of constituencies alternative to them. Powerful widely-held big business independent from politics could play this role of alternative constituency. Politics thus pursued institution building in order to curb their growth. By contrast, growth through the use of mechanisms that amplified financial leverage, such as BGs, was preferred. In fact, this created a situation in which the dominant shareholders must continuously bargain favourable regulation and access to resources with the political power (especially credit from a largely state-controlled banking system, as it was the case from the 1930s to the 1990s), thereby making them more dependent on politics. 65

Conclusion
In this article, we put forward a series of conceptual contributions in current research about BGs, looking at the Italian case in a historical perspective.
A first, important point is that, as the Italian case shows rather clearly, BGs are not an organisational form peculiar to developing countries or of economies in their very early stage of development. In the course of its evolution from a peripheral economy to one of the most industrialised countries of the world, the Italian corporate system has been constantly populated by BGs, dominant in their diffusion and economic relevance over other organisational structures.
Secondly, this article shows that BGs are present not only among large firms, but in almost all the dimensional categories and juridical forms. This challenges the conventional wisdom which assumed that BGs are simply an alternative to the M-form characterising big business around the world.
Third, this article demonstrates that vertical BGs, often identified with family ownership, can in fact be diffused also among non-family firms, ranging across a wide spectrum of ownership forms, as co-operatives, municipalised companies, and SOEs.
Our contribution suggests also how longitudinal analysis is essential to understand in-depth the whole phenomenon of the diffusion of BGs in a country. The BG as a governance system looks particularly flexible, adapting itself to different ownership and market conditions and local contexts. In the absence of obstacles of legal or fiscal nature, this flexibility is probably the main reason for its resilience.

Notes on Contributors
Andrea Colli, PhD in Economic and Social History (Bocconi university, Milan) and Professor of Economic History at the Department of Policy Analysis and Public Management, Bocconi university, Milan. His research interests range from the history of Italian business enterprises, to small and medium-sized enterprises, to the role played by foreign direct investments in the course of the Italian industrialization process. Among his most recent publications, Business History: Complexities and Comparisons, Routledge 2011, with Franco Amatori. Alberto Rinaldi has received his D.Phil at Bocconi university, Milan. Now he is lecturer of Economic history at the Department of Economics "Marco Biagi" of the university of Modena and Reggio Emilia. He has published extensively on Italian economic history, focusing in particular on industrial districts, trade, economic growth, and corporate networks. His works have been published in leading economic and business history journals such as Explorations in Economic History, Business History, and Enterprise and Society. Michelangelo Vasta has received his D.Phil at the university of Oxford. He is now professor of Economic history at the Department of Economics and Statistics of the university of Siena. Most of his research work has dealt with Italian economic development from the unification to the present. His fields of research range over macro and micro perspective and focus on technical change, institutions, corporate networks and international trade. He has been published extensively on the major economic and business history journals such as Economic History Review, European Review of Economic History, Explorations in Economic History, Business History and Enterprise and Society.