Jump to ContentJump to Main Navigation
The Spread of Modern Industry to the Periphery since 1871$

Kevin Hjortshøj O'Rourke and Jeffrey Gale Williamson

Print publication date: 2017

Print ISBN-13: 9780198753643

Published to Oxford Scholarship Online: March 2017

DOI: 10.1093/acprof:oso/9780198753643.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 02 December 2021

The Industrialization of Italy, 1861–1971

The Industrialization of Italy, 1861–1971

Chapter:
(p.115) 6 The Industrialization of Italy, 1861–1971
Source:
The Spread of Modern Industry to the Periphery since 1871
Author(s):

Matteo Gomellini

Gianni Toniolo

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198753643.003.0006

Abstract and Keywords

At unification in 1861, the Italian peninsula was a relatively backward area at the European periphery. By 1971, Italy’s convergence on Europe’s northwestern industrial core was almost complete. This chapter describes the main features of Italy’s industrial and manufacturing growth, emphasizing the role of traditional and modern sectors. It assesses the impact of commercial and industrial policies, and analyses the country’s regional manufacturing divide. The chapter concludes with a list of the main drivers of the spread of manufacturing over the long run. It raises the question of the timing of the spread of industry from core to Italian periphery. On the eve of the Second World War, the shares of modern sectors in manufacturing were close to those of core countries. However, while Italy’s Northwest looked like an industrial region, the South was still part of the backward periphery—a divide that reduced only moderately after the war.

Keywords:   Italy, manufacturing, economic geography, economic history, regional inequality, core, periphery

6.1 Introduction

Between 2007 and 2012, as a result of its longest (and possibly deepest) depression since the country’s unification (in 1861), Italy’s manufacturing output decreased by more than 21 per cent (stats.oecd.org).1 Even so, according to UNCTAD, in 2012 Italy was still the world’s sixth largest manufacturer, slightly ahead of the United Kingdom and France, each of the three countries accounting for about 3 per cent of total world manufacturing production. This is hardly surprising given Italy’s size and its natural endowments. Poor in raw materials and (outside the Po Valley) in agricultural land suitable for modern capitalist agriculture, Italy was bound to rely on manufacturing for its ‘modern economic growth’. In 2011, value added by industry still accounted for 18.6 per cent of GDP in Italy as against 16.2 in the USA, 14.9 in Great Britain, and 12.5 in France (OECD, 2013). In order to prosper, as Carlo Cipolla famously said, ‘Italy must manufacture goods that please the world.’

For three centuries (from around 1300 to 1600) Italy was at the core of Europe’s economic activity, particularly in manufacturing. Over the following centuries the country gradually slipped to the ‘periphery’. At the time of its political unification in 1861–70, its GDP per person was about 50 per cent of that of Great Britain. In the next century, Italy’s convergence with the most prosperous European countries was broadly as fast as might have been expected given its initial backwardness (Toniolo, 2013). By the 1980s GDP per person was roughly similar to that of Great Britain. The history of the Italian economy moving from periphery back to core is the subject of various works (e.g. Zamagni, 1993; Castronovo, 1995; Toniolo, 2013): this chapter focuses on the role played by industrial production, particularly manufacturing, in Italy’s catch-up growth.

The quantitative history of Italy’s industrialization, while not matching that of Britain and the United States, is as rich as, or richer than, that of most of today’s OECD countries, thanks to a nineteenth- and early-twentieth-century tradition of descriptive statistics and to the post-war work of the Statistical Office (ISTAT) and (p.116) individual scholars such as Gerschenkron, Fenoaltea, Ercolani, Vitali, and Carreras, to name only those most active in the field of industrial historical statistics. Industrial censuses were taken in 1911, 1921, and 1936–8. After the Second World War industrial statistics were routinely produced by ISTAT. Recently, estimates of industrial productivity growth have been extended back to the 1860s (Broadberry, Giordano, and Zollino, 2013). In this chapter, aggregate industrial value added is taken from the most recent reconstruction of national accounts (Baffigi, 2013);2 sectoral estimates are those by Ciccarelli and Fenoaltea (2009) for the period 1861–1913; by Carreras and Felice (2012) for the inter-war period; and by ISTAT and Golinelli and Monterastelli (1990) for the second post-war period.

The chapter is organized as follows. Section 6.2 traces aggregate trends in Italian manufacturing output and productivity growth from 1861 to the early 1970s, a time when Italy’s industrial system was no longer ‘peripheral’ by any definition. Section 6.3 focuses on ‘modern’ vs ‘traditional’ manufacturing sectors and on utilities. Section 6.4 discusses the impact on industrial output of foreign trade, tariffs, and industrial policy. Sections 6.5 and 6.6 develop a quantitative analysis of regional industrial growth. The final section pulls the main threads together to provide an overall assessment of Italy’s industrial growth from the 1860s to the 1960s.

6.2 The Aggregate Picture

According to Bairoch (1982), at the time of political unification (1861) Italy’s industrial output per person was less than one-sixth (16 per cent) of the UK’s, a level roughly similar to that of other peripheral West European countries (Spain, Denmark, Finland, and Norway) and higher than that of both Eastern Europe (Russia, Bulgaria, and Romania) and most non-European countries (including Japan, China, Mexico, South Africa, Australia, and New Zealand). On average Italy was less industrialized than Bairoch’s ‘developed countries’ and more so than his ‘third world’. At the same time, Italy’s industrial backwardness relative to the leading manufacturing power of the day reflected more than proportionally the country’s overall economic backwardness since, according to Maddison (2010), in 1870 Italy’s GDP per person was about 50 per cent of the UK’s level.

Italy’s economic history in the 150 years following unification has been told as ‘a tale of convergence and two tails’ (Toniolo, 2013). As expected, during this period, GDP and productivity of this ‘moderately backward country’ à la Gerschenkron converged to those of the initially most advanced countries. Convergence in GDP per person, however, took place over the years 1896–1995 while the early post-unification decades and the post-1995 years were characterized by divergence rather than convergence. This section describes Italy’s aggregate industrialization against the backdrop of its overall economic development.

(p.117) A feature of Italy’s industrialization is that its share of world industrial output remained roughly constant over more than two centuries: it was 2.5 per cent in 1750, remained at that level until the First World War, rose to 2.7 per cent in 1938, and reached about 3 per cent in 1973 (Bairoch, 1982). In contrast, the UK’s share of world industrial production was 1.9 per cent in 1750, peaked at 22.9 per cent in 1880, and declined thereafter to 4.9 per cent in 1973. A similar inverted U-shape curve characterizes Germany’s share of world industry and, to a lesser extent, France’s. In other words, Italy followed a pattern roughly similar to that of the overall international industrialization process.

Fig. 6.1 describes the weak unconditional convergence in industrial value added per capita realized in a sample of 30 countries during 1880–1973 (data from Bairoch, 1982).3 Italy performs roughly as expected given its initial level of backwardness, while initially less industrially developed European countries (e.g. Bulgaria and Russia) seem to converge to the ‘core’ more rapidly than Italy (and Spain and Portugal).

The Industrialization of Italy, 1861–1971

Fig. 6.1. Unconditional convergence of industrial value added (all available countries)

It is worth noting that, while industrial value added grew faster in Bulgaria and the former Soviet Union than in Italy, these three countries did not significantly catch up with Italy in terms of GDP per person, indicating that Italy’s ‘modern economic growth’ involved more than just industrialization.

(p.118) Italy’s manufacturing history between 1861 and 1973 can be divided into five sub-periods (Table 6.1): (i) relatively slow output growth (1861–96), (ii) growth acceleration (1896–1913), (iii) rapid output and labour productivity growth (1922–9), (iv) slow growth, protection, currency revaluation, great depression, sanctions, and autarky (1929–39), (v) a long wave of catch-up growth (1951–73). The two world wars can both be characterized as ‘lost decades’, albeit involving very different stories and impacts on long-term manufacturing growth.

Table 6.1. Average annual growth rates of GDP and manufacturing

GDP

Manufacturing value added

Manufacturing value added per full-time equivalent worker

1861–96

1.7

2.0

2.0

1896–1913

2.2

3.6

2.1

1922–9

4.0

6.3

4.2

1929–39

1.5

2.1

−0.1

1951–70

5.9

8.8

6.0

Sources: Baffigi (2013); Broadberry, Giordano, and Zollino (2013).

6.2.1 Before the First World War

According to Malanima (2007), Italy’s industrial output and labour productivity gently declined from the beginning of the fourteenth century to the early nineteenth century when they began to pick up. At the time of unification a moderate upward trend was already detectible. Between 1861 and 1896 manufacturing value added grew on average by about 2.0 per cent per annum. Employment in the sector remained roughly constant, however, while the economy as a whole created 1.5 million new jobs. In 1896–1913 the growth rate of manufacturing accelerated to 3.6 per cent per annum. The sector’s workforce increased by about 1 million (growing by 1.5 per cent per annum as against 1.0 per cent for the economy as whole). Between unification and the First World War, output per full-time equivalent worker in manufacturing increased every year by 2.0 per cent (Broadberry, Giordano, and Zollino, 2013): a respectable rate by nineteenth-century international standards.

The pattern of Italy’s industrialization prior to 1914, characterized by slow growth up to the 1880s and growth acceleration thereafter, was the subject of a heated debate sparked in the late 1950s by Alexander Gerschenkron’s (1962) attempt to fit the Italian case into his paradigm of European industrialization in conditions of moderate backwardness, in which ‘agents’, such as universal banks (created in Italy in the mid-1890s), are ‘needed’ to overcome the ‘stumbling blocks’ that obstruct the path to modern industrial growth. Rosario Romeo (1959) argued instead that the sluggish growth in the first decades after unification was due to the ‘need’ for social overhead capital to be created before rapid industrialization could take place. The two scholars had a lively debate in Rome, later published by Gerschenkron (1962): Romeo spotted an industrial acceleration in the 1880s, (p.119) upon completion of the main rail network, while Gerschenkron dated Italy’s ‘big spurt’ a decade later, coinciding with the appearance of German-type universal banks. Bonelli (1978) and Cafagna (1972), on the other hand, saw Italy’s early industrialization as characterized by slow trend acceleration in industrial output, beginning before unification, rather than discontinuity at the end of the century.

In the late 1960s, Stefano Fenoaltea began to produce a new set of industrial value added statistics, which he continuously refined over the following decades. During the course of this exemplary work, Fenoaltea produced a new interpretation of Italy’s industrialization pattern, which he saw as driven by the flow of international investments and therefore following the international business cycle (Fenoaltea, 2006). Recently, the creation of new data on the engineering sector led Fenoaltea (2014) to doubt his previous assessment of Italy’s industrialization path, arguing, however, that additional work is needed to produce a new explanation.

6.2.2 The Great War and Inter-war Period

If we accept the most recent estimates (Baffigi, 2013), wartime manufacturing output grew during 1914–16 by 5 per cent per annum. (Italy entered the war in May 1915). Thereafter the war-related acceleration was not sustained, and manufacturing value added fell during the second part of the conflict and the subsequent recession. In 1922, manufacturing value added was not significantly different from 1914. The war, however, marked a qualitative turning point in Italy’s industrialization, accelerating technical progress.

A period of exceptionally rapid manufacturing growth followed the post-war depression and industrial restructuring. During 1922–5, the sector grew at an average annual rate close to 8 per cent. In spite of the subsequent slowdown, the 1920s saw a remarkable performance of Italy’s manufacturing, with high rates of output, productivity, and employment growth.

The 1930s can be divided into two periods: the Great Depression and the autarkic recovery. New estimates (Giugliano, 2011; Giordano and Giugliano, 2012; Carreras and Felice, 2012; Baffigi, 2013) indicate that the fall in industrial value added and employment during 1929–32 was deeper than previously believed: in three years, manufacturing output fell by 22 per cent. Recovery was slow in 1933–4 and rapid thereafter, partly driven by war-related (Abyssinian and Spanish) expenditure. In 1939 (Italy entered the conflict in June 1940), manufacturing value added was 45 per cent higher than in 1934. On the whole, Italian manufacturing was highly volatile during the inter-war years, with spurts of high growth followed by sharp slowdowns and a slump during the Great Depression. As we shall see, economic policy played an important role alongside the international business cycle.

Unlike the First, the Second World War had a strong negative impact on Italy’s manufacturing production, which shrank to a quarter of the pre-war level. The decline started immediately after the beginning of the hostilities, remained relatively contained in 1941–2, and accelerated during 1943–5 when war was waged on Italian soil, and the country was divided into two separate states at war with each other.

(p.120) 6.2.3 1945–73

As in other defeated countries, the reconstruction of Italian industry was remarkably swift. In 1946, despite shortages of raw materials and the disruption of transport infrastructure, industrial value added grew by 242 per cent. The pre-war peak (1940) in manufacturing was reached in mid-1950.

In the twenty-odd years following reconstruction, cyclical stability stood in sharp contrast to inter-war volatility. The overall picture can be summarized by saying that in 1970 Italy’s manufacturing sector was almost five times bigger than it had been in 1950, having grown on average by 8.3 per cent over the two decades. Between 1950 and 1958 manufacturing value added grew on average by 7.6 per cent per annum, accelerating to 10.7 per cent between 1958 and 1963. After a brief slowdown to just 2.4 per cent in 1964, growth resumed at the annual pace of 8.6 per cent until the end of the decade.

To conclude this section with a comparative perspective, Table 6.2 reports the shares of industrial employment in Italy and five other countries between 1871 and 1973. The comparison with Japan is particularly striking.

Table 6.2. Industriala employment as a percentage of total employment

Yearb

Italy

UK

Germany

US

India

Japan

1871

15.8

42.4

29.1

24.8

12.1

10.5

1911

23.5

44.1

37.9

31.8

10.3

1921

22.5

33.2

19.0

1931

24.5

43.7

37.4

30.2

9.1

1936

25.6

44.5

38.2

1951

31.1

46.5

42.1

32.9

10.2

22.9

1973

38.4

41.8

47.3

28.9

11.1

36.6

a Including construction

b Italy; for some other countries, the closest available year.

Source: Broadberry et al. (2013).

6.3 ‘Modern’ and ‘Traditional’ Sectors

In the beginning was silk. So the story goes, as told by a leading Italian economic historian (Cafagna, 1972). Raw silk manufacturing had a number of advantages for a relatively backward agricultural country. The cultivation of mulberry trees, for the production of cocoons, and silk spinning were useful labour-intensive complementary activities for under-employed agricultural workers, particularly in hilly areas where land productivity was not as high as at the bottom of the Po Valley. Furthermore, Federico (2005) finds strong evidence of a beneficial influence of raw silk production on engineering: already at the time of unification, Italy was the world leader in the production of silk-reeling machinery, a leadership it kept until the 1920s (Federico, 1997, 2005). In the decade before unification, a silkworm disease had considerably reduced the production of cocoons (Federico, 1997), sending the raw material price to (p.121) levels only affordable by firms equipped with steam-powered machinery (Federico and Tena, 1998). Initially, boilers were produced by non-specialized engineering firms but, as technology became more sophisticated, a handful of new specialized producers emerged as the result of growing demand and weak foreign competition. A similar pattern of demand-driven production of investment goods would become common after the Second World War in the so-called industrial districts, which often developed the technology for the machinery used in the production of consumer goods. In the years immediately following unification, large-scale investment in steam-reeling resulted in the silk sector’s total factor productivity growth being 3.5 times faster than the economy’s average.

If silk figures prominently in the narrative of Italy’s early industrialization, one should not exaggerate its importance: in 1870, the total production of textiles accounted for only 10 per cent of manufacturing value added, while engineering made up 17 per cent (Fenoaltea, 2006). If industrialization is to be the main factor in the process of ‘modern economic growth’, as described by Simon Kuznets, then it should rely upon industries that generate more innovation, investment, R & D, and productivity growth—in a word ‘modernization’—than silk. For our purposes, considering that comparable time series for only thirteen industrial sectors are available for the entire period under review,4 we take metallurgy, engineering, and chemical products (including rubber) to make up the bulk of the ‘modern sector’ (Table 6.3). This is admittedly a very rough approximation for at least two reasons: first, each sector’s innovation, technical progress and R&D varied over time;5 second, seeds of modernity, new products and new production processes abounded in other sectors (a typical example is the rapid development of viscose rayon (so-called ‘artificial silk’) within the ‘traditional’ textile sector during the 1920s).

Table 6.3. Shares and annual average growth rates of the ‘modern’ sector,a 1870–1973

‘Modern’ sector’s share in manufacturing

Average growth in ‘modern’ sectors

Average growth in manufacturing

Average growth in utilities

Average growth in electric power production

1870

0.21

1890

0.27

1870–90

3.3

2.1

5.1

n.a.

1913

0.29

1890–1913

3.8

3.0

8.3

27.7

1920

0.28

1913–20

11.4

1925

0.35

1920–5

12.4

8.0

9.6

9.1

1929

0.37

1925–9

8.5

1.6

8.6

9.3

1939

0.47

1929–39

10.3

1.5

6.4

5.7

1951

0.45

1973

0.56

1951–73

8.6

7.3

7.8b

7.6

(a) Metal making, engineering, and chemicals (incl. rubber).

(b) Energy only.

Sources: 1870–1913, Ciccarelli and Fenoaltea (2009); 1920–39, Carreras and Felice (2012); 1951–73, Golinelli and Monterastelli (1990).

The overall pattern of sectoral growth that emerges from Table 6.36 is pretty standard in any industrialization process: an increasing share of ‘modern sectors’ within manufacturing accompanied a growing share of manufacturing in GDP. More specific to Italy is perhaps the very rapid expansion of hydroelectric power production, which grew from almost nothing (8 million Kw) in 1890 to 2.2 billion Kw in 1913, more than doubled during the Great War (1913–20), and grew thereafter at a slower pace, close to that of the overall utilities sector.

More interesting than Italy’s ‘normal’ industrial growth pattern over the very long run is perhaps what emerges from the observation of individual sub-periods. Before the First World War, ‘modern sectors’ made a relatively modest contribution to manufacturing growth. In half a century, the ‘modern’ share in manufacturing increased from only a fifth to less than a third of the total. On the other hand, as mentioned above, from the end of the nineteenth century to the early post-war (p.122) period, electrification proceeded rapidly, liberally licensed by the state and financed by equity issues and long-term loans from universal banks, in response to both industrial and civil demand.

The inter-war period stands in striking contrast to the pre-war years. In 1920–5, both modern and traditional industry grew at a rapid pace, while after 1925, non-modern manufacturing almost stagnated. As discussed in the next section, tariff protection and possibly the overvaluation of the currency were instrumental in shifting resources from the labour-intensive, export-oriented sectors to the more capital-intensive ones so that, in 1939, the ‘modern sector’ share in manufacturing was close to 50 per cent. From this viewpoint, a case can be made that, by that time, the structure of Italian industry was more similar to that of ‘core’ than ‘peripheral’ countries. The industrialization pattern during the post-Second World War golden age looks similar to that of other countries (including ‘core’ ones such as Germany, France, and Japan). Average yearly growth for our ‘modern’ sectors outperforms the average for total manufacturing by about 1.5 percentage points, indicating that value added (and probably productivity) growth was by no means limited to the capital-intensive ‘modern’ industries. In fact, our definition of ‘modern’ itself might no longer apply to the latter part of the period.

6.4 Trade, Tariffs, and Industrial Policy

Among the factors affecting a latecomer country’s manufacturing performance and catch-up, the most frequently highlighted in the literature are comparative advantages and terms of trade, tariff and industrial policies, and the quantity and direction of foreign capital inflows (Blattman, Hwang, and Williamson, 2007; (p.123) Williamson, 2008, 2011). In what follows we briefly discuss the impact of these factors on Italy’s industrialization.

Since unification in 1861–70, Italy has always been a medium-sized economy, one of the eight to ten largest in the world. Not nearly as large as those of the United States, China, or the German and British Empires, Italy’s domestic market was nevertheless large enough to provide potential economies of scale for ‘modern’ industrial sectors. Italy’s openness ratios (trade/GDP) rose from 0.2 to close to 0.3 in the second half of the nineteenth century, declined to 0.1 in the 1930s, and rose to about 0.4 in the 1970s. Throughout our period Italy’s openness was lower than Germany’s, higher than Spain’s, and close to that of France (Federico and Wolf, 2013). Given the medium size of the economy, the choices between free trade versus protectionism, or fixed versus floating exchange rates, were less straightforward than in smaller countries. Industrial policy, which mainly took the form of the state directly or indirectly allocating resources to specific manufacturing sectors, could count on relatively large state budgets and domestic capital markets. In a word: size mattered in shaping policy options.

Touring Italy soon after unification, Richard Cobden advised the new Italian rulers to capitalize on Italy’s comparative advantages (climate and art), leaving manufacturing to countries north of the Alps (possibly north of the Channel). At the time, Cobden’s recommendation might not have been far off the mark. When measured by the Lafay (1992) index, a commodity-specific net exports measure of comparative advantage, until the turn of the twentieth century Italy did not enjoy a comparative advantage in manufacturing when raw silk production is excluded (Federico and Wolf, 2013, pp. 239–40). However, Lafay’s index moves from −10 to 0 between the 1880s and the late 1890s, indicating that Italy was ‘building’ a manufacturing comparative advantage. Poor in coal, iron ore, and oil, Italy did not suffer from the ‘curse of raw materials’: sulphur, tomatoes, olive oil, and citrus fruits were not sufficient to develop a serious curse.

Throughout the twentieth century, by Lafay’s measure, Italy had a clear comparative advantage in manufacturing production. The share of manufacturing in Italian exports rose from only 20 per cent in the 1870s and 1880s to almost 40 per cent in 1914, 50 per cent in the late 1930s (including colonial trade), and about 80 per cent in the mid-1970s. Following a standard pattern in most industrialization processes, Italy’s exports, initially concentrated in a few products, became progressively diversified.

After the Second World War, Italy’s share of world exports increased considerably. The composition of manufacturing exports changed, without however fully converging on that of the more advanced countries (Gomellini, 2004; Gomellini and Pianta, 2007). In particular, the country’s export success was to a large extent based on goods of mid-innovative content, produced mainly by medium-sized firms. At the end of the 1960s, Italy’s manufacturing still lagged behind in science-based and scale-intensive production, while exhibiting a leadership in more traditional manufacturing industrial products. The specialization in mid-tech productions that emerged in the 1960s persisted in the following decades.

(p.124) Italian economic historians have largely neglected the terms of trade, possibly reflecting the belief that they did not matter much one way or the other for GDP and manufacturing growth. A recent quantitative analysis covering the period from unification to the eve of the Second World War concludes that trade diversification ‘improved the Italian terms of trade but did not reduce volatility’ (Federico and Vasta, 2010, p. 236). Scarcely endowed with basic industrial raw materials, Italy was exposed to the vagaries of international prices for basic commodities, for which the country was an international price taker. There are no estimates available of the impact of terms of trade volatility on the rate of growth of manufacturing; a likely guess is that it had a very moderate negative effect.

Italy’s tariff history follows the European continental pattern. Born a free trader in the political and intellectual climate of the Cobden–Chevalier treaty, the Italian state leaned towards a more protectionist mood from the late 1870s onward, in tune with the main continental countries. In an effective protection framework, Gerschenkron (1955) argued that the 1887 tariff was detrimental to industrial growth because of its protection for wheat growers and the high duty on iron and steel—inputs to the promising engineering sector. Toniolo (1977) concurred but argued that the impact of the steel duty on the rate of growth of engineering was empirically modest. Other scholars (Romeo, 1959, 1988; Zamagni, 1993) were more sanguine about the tariff of 1887, invoking infant industry arguments and approving of the active involvement of the government in the promotion of industry. More recently, Fenoaltea—the leading scholar of pre-1914 industrialization—showed that the 1887 tariff and the subsequent periodic increases in the duty on wheat had a non-negligible negative impact on the growth rate of Italy’s manufacturing (Fenoaltea, 2006, Chapter 4). Federico and Tena (1998) estimate that the tariff of 1887 brought the average rate of Italian protection to 50 per cent above the European average. However, while manufactured goods accounted for about 32 per cent of all imports, they accounted for only 20 per cent of the tariff revenue increase. The nineteenth-century debate on tariffs and growth went on to the present day: the prevailing view is that protection—of both agricultural and industrial goods—slowed Italy’s manufacturing growth before the First World War, but the size of the damage done by the tariff is still being debated.

The 1887 tariff was introduced when a slowdown in GDP and manufacturing growth was beginning to take place. From the end of the nineteenth century to the outbreak of the Great War, industrial protection was consistently eroded by the signing of commercial treaties (Toniolo, 1990) and by a decline in the undervaluation of the real exchange rate (Di Nino, Eichengreen, and Sbracia, 2013). The period coincides, as we have seen, with an acceleration of industrial growth. By 1914 Italy’s average manufacturing tariff (estimated at 18 per cent) stood at a level close to Germany’s and France’s, lower than Spain’s and Sweden’s, but much higher than that of small open economies such as Belgium, the Netherlands, and Switzerland (James and O’Rourke, 2013, p. 41).

A new protectionist turning point took place in 1925: high import duties on wheat and sugar were re-introduced, and metallurgy and a number of heavy (p.125) engineering activities regained the customs’ favour. A discussion of the underlying political economy of the protectionist U-turn of 1925 is beyond the scope of this chapter,7 but it is enough to recall that 1925 was the year when Mussolini’s dictatorial powers were sanctioned by law.

Italy’s protectionist turn antedates by a few years the international triumph of tariffs, quotas, and exchange controls in which the country fully participated, first proclaiming a national ‘battle for (domestic) wheat’ and later embarking on an official programme of autarky. The latter relied on an array of allocative tools, among which import duties were just one of many.

Post-1945 tariff history closely followed the European pattern, first in the General Agreement on Tariffs and Trade (GATT), and then in the European Coal and Steel Community (ECSC) and European Economic Community (EEC) frameworks. The swift reconstruction of the Italian industrial base took place under a heavy protective shield, which included import duties, quotas, restrictions on capital movements, cumbersome authorization procedures, and managed exchange rates. The panoply of protective devices began to unwind in the late 1940s, following the Annecy conference, when Italy joined the GATT. In 1950, a tarif de combat was introduced, in the time-honoured tradition of entering into negotiations from a high ground. Progressive liberalization followed and, by 1957, Italy was ready to join the European Common Market and to make its currency convertible. From then onward, Italy’s import duties were set in Brussels as trade liberalization in manufacturing, both among the six original members of the EEC and with the outside world, proceeded faster than was originally agreed upon in the Rome treaty.

To sum up: Italy’s tariff history followed the ebbs and flows of European trade policy. Both the 1887 and 1925 duties on manufacturing were aimed at shielding so-called heavy industry from foreign competition. The former also protected cotton textiles, which were neglected in 1925. The prevailing consensus among scholars is that in both cases tariffs had a negative impact on manufacturing growth which, on the other hand, was enhanced by post-1945 trade liberalization.

When not raised mainly for fiscal purposes, which was never the case in Italy, tariffs serve two political economy purposes: consensus building (as with the notorious pactum sceleris between agrarian and industrial constituencies in Germany, but also in Italy) and ‘industrial policy’. The latter is a broad and often ill-defined concept: for our purposes we define ‘industrial policy’ as the use of the state’s power to influence resource allocation in order to promote industrialization (wartime policies being the extreme case in point). For better or worse, industrial policy looms large in Italy’s industrialization drive. The tariff hikes of the 1880s, and 1920s and 1930s, were both part of broader policy drives aimed at promoting ‘strategic’ industries.

The (liberal) left that came to power in 1878 had a more hands-on attitude to industrialization than the ‘Historical Right’, which had been in government since (p.126) the Risorgimento.8 Besides the tariff, a number of industrial policy tools were used: ‘ad hoc legislation, procurement, grants, fiscal privileges, bail outs, state guarantees […] all administered with large discretion, and to changing ends’ (Ciocca, 2007, pp. 116–17). Lack of transparency resulted in the exchange of favours between the private and the public sectors, cutting corners, and careerism, all captured by the novels of Maupassant in France9 and De Roberto in Italy.10

Metal making saw double-digit growth in the 1880s (12 per cent per annum, up from 2.9 in 1861–80). Engineering, already the second largest manufacturing sector after food processing, grew annually by 5.3 per cent (up from 2.3). But the textile sector, not a target of government activism, also grew faster than the manufacturing sector on average. How much of the manufacturing growth acceleration can be attributed to the new activism by the left government? Probably not much of it, if other favourable factors are taken into account. Major institutional progress had been made in monetary and financial market unification (Toniolo, Conte, and Vecchi, 2003), and in the creation of business-friendly institutions (notably the Commercial Code of 1882). Foreign capital flowed in (Fenoaltea, 1988) as the result of a balanced state budget and the return to gold convertibility of the lira in 1883. A construction boom (4.5 per cent per annum, up from a negligible and volatile 0.36 per cent per annum during 1861–80) created considerable private demand for iron and steel, unrelated to government policy.

All in all, the government’s chaotic panoply of support to manufacturing either slowed down growth, as maintained by economists at the time, or contributed little to it. Moreover, the brief cyclical expansion (1880–8) was followed by a slump in 1888–96, during which manufacturing grew by a meagre 0.84 per cent per annum while metallurgy, engineering, and textiles, as well as construction, all saw negative growth rates. If industrial policy is to be judged by its long-term impact, the left’s activism does not pass the test.

As we have seen, during 1896–1925 Italy made a major quantitative and qualitative jump in modern industrialization, narrowing the gap with the core countries. Did industrial policy matter?

Giolitti, who directly or indirectly ruled the country from the end of the century to 1913, did not put an end to previous practices of the government paying excessively close attention to the needs of individual sectors and even of particular business groups. An active industrial policy, however, began to be conducted by the universal banks.

(p.127) It is impossible here to even review the vast literature on the contribution of German-type mixed banks to the manufacturing growth acceleration in Italy after the mid-1890s. The prevailing consensus (Confalonieri, 1994) is that, while Banca Commerciale and Credito Italiano introduced innovative banking practices and played an important role in the development of some sectors (notably hydropower), in underwriting equity and bond issues, and more generally in developing the financial market, they did not play the pivotal role as ‘agents of industrialization’ (i.e. as industrial policy-makers par excellence) assigned to them by Gerschenkron. The most important developmental role of the government during the Giolittian period was a time-consistent monetary and fiscal policy, which stabilized expectations by both domestic and foreign investors, lowered interest rates, crowded in private capital, and favoured a hefty inflow of emigrant remittances.

In every war economy, industrial policy is pushed to the extreme. Every tool available to the state—selective credit and taxation, price controls, licensing, tariffs, labour laws, and many others—is used to allocate resources to the production of war-related goods. Policy is enforced by special legislation, including the extension to civilians of the tough military penal code. For all its deficiencies and shortcomings, both military and industrial, Italy’s wartime economic performance was quite satisfactory, given the relative backwardness of the economy, the inefficient public administration, and the social and political fractures that had emerged in the previous years. As far as industrialization is concerned, the First World War (unlike the Second) generated leaps forward in technical progress. There were two main drivers: economies of scale and import substitution. As for the former, the huge government demand for arms, ships, vehicles, and ammunition led to the creation or expansion of conglomerates, such as Ansaldo, for the vertically integrated production of steel, engineering products, warships, guns, trucks, and railway equipment. Another instance can be found in the wartime expansion of FIAT, the main producer of transport equipment. Import substitution was particularly relevant in fostering growth in the chemical industry, previously largely dependent on German imports. Montecatini developed the production of such products as superphosphates, nitrogen, synthetic colours, and sodium carbonate for the preparation of caustic soda. In several cases these products were inputs to other industrial processes, particularly explosives (Caracciolo, 1969; Galassi and Harrison, 2005). Some of the largest companies were downsized in the transition to the peace economy, and excess capacity emerged in some sectors (notably shipbuilding), but most of the qualitative changes to manufacturing induced by wartime ‘industrial policy’ were not lost in the transition. In particular, as engineers and skilled industrial workers had been largely exempted from serving in the trenches, the wartime manufacturing effort resulted in a permanent increase in the country’s human capital through learning by doing and tacit knowledge.

The next relevant chapter in industrial policy was inaugurated by the protectionist U-turn of 1925, which favoured capital-intensive sectors (particularly steel making, engineering, chemicals, and sugar refining). Besides import duties, the usual allocative tools were employed: procurement, long-term credit at favourable conditions, grants and state guarantees. The revaluation of the lira, which started in (p.128) the summer of 1926, was part of the new policy orientation. It helped in attracting foreign capital and, even more, in aiding large companies to repay foreign debts on favourable terms. Contrary to a widespread belief both at the time and among historians, the nominal revaluation had a relatively minor impact on the real exchange rate as it was accompanied by tight wage controls, and even by ‘wage cuts’ imposed by decree (Toniolo, 1980; Di Nino, Eichengreen, and Sbracia, 2013). The losers in this game were labour-intensive, export-oriented industries, which had led the post-war industrial boom, among them the rapidly expanding ‘artificial silk’ (rayon) sector in which Italy had a worldwide leading position as the second largest producer after the US.

The inward-looking nature of industrial policy was progressively reinforced in the 1930s. Administrative controls on capital movements and bilateral clearing agreements were added to the panoply as particularly effective allocative devices. An official autarky programme was eventually launched in 1935 as a reaction to the sanctions imposed by the League of Nations when Italy invaded Ethiopia.

The large number of state-owned enterprises (SOEs) is one of the most distinctive features of Italy’s industrialization. SOEs originated in the inter-war years mostly as the unintended consequence of government bailouts. In the 1900s, the government nationalized the railway companies, merging them into a single state-owned company, as well as life insurance, but no SOEs existed in manufacturing before the First World War. Government intervention to ease the transition from a war- to peace-time economy in the socially explosive conditions of 1919–22 resulted in the state gaining indirect control of Ansaldo, a large metal-making and engineering conglomerate. In 1931 the government secretly bailed out the three largest universal banks of the country which, in the 1920s, had progressively turned themselves into holding companies: their combined equity portfolios held controlling stakes in about 50 per cent of the listed companies on the Italian stock exchange. As a condition for the supply of last resort liquidity, the government required a commitment by the banks to henceforth confine their business to short-term lending; holding of industrial stakes was explicitly forbidden. In 1933, the equity portfolio of the banks was ‘provisionally’ taken over by a state-controlled holding company, the Istituto per la Ricostruzione Industriale (IRI).11 As far as ‘industrial policy’ was concerned, IRI coordinated and rationalized production within the group by directing financial resources where it deemed useful as well as by mergers and downsizing, the latter with little social cost as manpower could to a certain extent be reallocated within the group. From the very beginning, IRI also took responsibility for managerial training and selection for the entire group; a role efficiently performed, which particularly yielded fruit in the 1940s and 1950s.

During the post-war golden age of manufacturing growth, as tariffs rapidly lost importance, two main tools were used in industrial policy: credit allocation and public investment in manufacturing, largely but not exclusively channelled by SOEs. Active credit policies took place in the context of financial repression that (p.129) characterized Italy and the rest of Europe from the 1930s through the 1970s and beyond. Under such conditions, Battilossi, Gigliobianco, and Marinelli (2013) argue that during 1948–70 Italian banks effectively promoted growth in the Italian economy. Co-integration analysis ‘shows that the volume of credit to the industrial sectors tended to adjust to changes in the growth opportunities [of individual industries] as proxied by price/earning rates’ (Battilossi, Gigliobianco, and Marinelli, 2013, p. 513). Credit allocation under financial repression was possibly not particularly detrimental to industrial growth. In the golden age, SOEs made a huge contribution to industrial investment, produced considerable R & D, and are credited with creating positive externalities for the private sector. ‘IRI helped the development of the machinery industry by providing cheap intermediate inputs, compensating for the weakness of private firms in the field’ (Crafts and Magnani, 2013, p. 80). It also created an extended network of superhighways (autostrade), which lowered transportation costs and favoured the rapid expansion of the car industry. ENI developed gas fields in the Po valley and freed the country’s oil supply from the ‘Seven Sisters’ oligopoly.12 Between 1958 and 1969 IRI’s investments grew on average by 8.7 per cent per annum, and employment by 3.2 per cent (to 295,000), producing profits year after year (Ciocca, 2014, p. 147).

6.5 The Regional Divide

An outstanding feature of Italy’s modern economic growth is the stubborn persistence of regional inequalities in GDP per person. Whereas the country as a whole converged with the initially most advanced countries, a similar process did not take place within the country itself. Regional divides are observed in almost every country, but nowhere else is the phenomenon as deep and persistent as in Italy. Moreover, in contrast to other countries such as Spain, Italy is characterized by a persistent geographical pattern of inequality, with a ‘continuing dominance of one area of the country’ (see A’Hearn and Venables, 2013, p. 626).

An enormous literature exists on the ‘Southern Question’ (Questione Meridionale) dating back to the late nineteenth century.13 In recent years, various estimates of regional GDP from 1861 to the present have produced some consensus about the evolution of the regional income gap. However, regional historical industrial statistics are still under-researched, with the exception of the painstaking analysis (p.130) by Ciccarelli and Fenoaltea (2009, 2014), which unfortunately covers only 1861–1913.

In this chapter we begin to bridge the statistical gap. To do so, we rely for 1861–1913 on the series by Ciccarelli and Fenoaltea (2009).14 For subsequent years we follow Fenoaltea’s methodology by attributing to each region, at census benchmark years, a share of the national industrial and manufacturing production equal to the share of the regional labour force in the two sectors, using data produced by Vitali (1970).15 This yields regional industrial and manufacturing value added at constant prices for sixteen Italian regions at benchmark years from 1871 to 1971.16

In Table 6.4 regional per capita industrial value added is shown in Geary–Khamis 1990 dollars to allow for comparisons with a selected number of countries. It shows that the Northwest, comparable in population size to Sweden, was as industrialized as the Scandinavian country by the 1930s, while the other Italian macro regions lagged behind most of Europe’s countries.

Table 6.4. Industrial value added per inhabitant: Italy, four Italian macro areas, and selected countries, 1891–1971 (Geary–Khamis 1990 dollars)

Year

ITA

NW*

NE*

CEN*

S*

FR

SP

JAP

BEL

FIN

SW

NL

1891

332

402

278

295

256

968

484

137

1,440

282

341

1,070

1911

497

761

472

380

391

1,231

555

287

1,637

491

719

1,337

1931

647

1,047

494

569

396

1,676

703

574

1,965

631

995

1938

653

1,273

695

675

520

1,486

809

1,995

1,076

1,310

1961

1980

4,102

2,065

1,884

1,009

2,484

2,942

2,532

3,022

1971

3,515

5,869

3,596

3,023

1,919

3,991

2,581

3,423

4,395

3,777

3,840

4,607

Δ‎% 1891–1971

3.0

3.3

3.2

3.0

2.6

(*) NW = Northwest; NE = Northeast; CEN = Centre; S = South.

Source: Regional GDP/per person in Italy, Felice and Vecchi (2013); Regional Industry/GDP, Table 6. Other countries: GDP per capita: Maddison (2010); Ind./GDP, Mitchell (2007); Ind./GDP France 1961 and 1971 INSEE; 1971 data for Spain, Japan, and the Netherland are from Groeningen Growth Centre (http://ggdc.net/).

The geographical pattern of Italy’s industrialization can be summarized as follows: a large gap in per capita industrial value added between the Northwest and the rest of the country already existed in 1891; during 1891–1938 the South lagged behind in industrial growth per capita, while the Northeast and Centre progressed very similarly, losing some ground to the Northwest; the Second World War deepened the North–South divide while during 1951–71 all three other macro areas caught up with the Northwest.

A striking feature of Italy’s industrialization is the concentration of manufacturing in the Northwest: in 1971, with about 27 per cent of Italy’s population, the region produced more than half of the entire country’s manufacturing value added. Note that the North–South divide in manufacturing is more pronounced than in industry as a whole. The Southern divergence began in 1911, when the area produced about one-third of the country’s manufacturing value added, a share that dropped to one-eighth sixty years later. In 1891 manufacturing made up about 80 per cent of total industrial production in all four macro areas. A century later, the manufacturing/industry ratio was still 0.8 in the Northwest, but had fallen to around 0.7 in the Northeast and Centre, and had plummeted to 50 per cent in the South. One of the reasons for the weak Southern industrialization is its disproportionate reliance on construction and utilities rather than manufacturing.

A similar pattern is revealed when we consider the spatial distribution of the ‘modern’ manufacturing sector, defined, as above, by the chemical, engineering, and metal-making industries (Table 6.5). Up to the end of the nineteenth century, (p.131) (p.132) over one-third of ‘modern’ industries were located in the South, with the Northeast and Centre lagging behind. The Southern share had more than halved by 1951. By the 1970s, over 50 per cent of manufacturing produced in the Northwest was in the ‘modern’ sectors, which accounted for two-thirds of the total ‘modern’ production of the country.

Table 6.5. Distribution of value added by modern sectors across macro areas

NWa

NEa

CENa

Sa

Total

1871

31.4

17.4

16.5

34.6

100

1881

35.8

16.6

15.4

32.2

100

1891

38.0

15.8

14.9

31.3

100

1901

40.7

14.9

14.4

30.0

100

1911

45.5

15.9

14.1

24.4

100

1921

49.5

14.4

13.4

22.7

100

1931

53.7

14.1

14.9

17.4

100

1936

52.8

15.0

14.5

17.7

100

1951

59.2

13.9

12.8

14.2

100

a NW = Northwest; NE = Northeast; CEN = Centre; S = South.

Sources: Our calculations; see text for data sources.

The rapid concentration of the modern sectors in the Northwest began with the post-1896 industrial growth acceleration, and was favoured by the wartime demand for typically modern products and by post-1925 tariff and industrial policies. After the Second World War, large enterprises remained the engine of growth in the Northwest, while the Northeast and Central regions developed their own idiosyncratic industrialization pattern based on small and medium-size enterprises, often clustered in ‘industrial districts’, typically—if by no means exclusively—oriented to the production of medium- rather than high-tech goods. SOEs were largely responsible for the development of modern sectors in the South.

By focusing on just four macro areas, one misses at least one important feature of Italy’s industrial (and general economic) growth, namely the large within-area dispersion of income levels. Fig. 6.2 shows modest convergence in industrial value added per person (upper graph) and divergence in income per capita (lower graph) for the sixteen original (1861) Italian regions. The scatter plots highlight the wide regional dispersion in both initial values and growth.

The Industrialization of Italy, 1861–1971

Fig. 6.2. Convergence in industry, divergence in GDP, 1871–1961

Note: Abbreviations refer to the sixteen Italian regions: ABR = Abruzzi; BAS = Basilicata; CAL = Calabria; CAM = Campania; EMI = Emilia; LAZ = Lazio; LIG = Liguria; LOM = Lombardia; MAR = Marche; PIE = Piemonte; PUG = Puglia; SAR = Sardegna; SIC = Sicilia; TOS = Toscana; UMB = Umbria; VEN = Veneto.

Regional divergence in per capita GDP (and in per capita manufacturing, as we shall see in the next section) is a well-known feature of Italy’s modern economic growth. Little attention, however, has been paid so far to the difference between the industrial and GDP stories, which suggests that industrial growth was probably a necessary but not sufficient condition for regional GDP convergence. The quality of the industrial mix (e.g. manufacturing vs construction, modern vs traditional sectors) mattered. (p.133)

(p.134) 6.6 Drivers of the Uneven Regional Spread of Manufacturing

As already mentioned, regional differences in per capita manufacturing output widened over the period under consideration (Fig. 6.3). In 1871, per capita value added by manufacturing in the South was about 70 per cent of the Northwest’s; in 1961 the figure was just 30 per cent.

The Industrialization of Italy, 1861–1971

Fig. 6.3. Divergence in per capita manufacturing, 1871–1961

Note: Abbreviations as in Fig. 6.2.

What drove this divergence? Deep history, government policies, institutions, geography, and human and social capital have all featured in the debate on the causes of the uneven spread of modern industry in Italy since the late nineteenth century.17

Galasso (2005) argued that Frederic II’s centralization of state functions weakened the autonomy of the Southern cities, which could not match the industrial vitality of their Central and Northern counterparts: a weakness that was not redressed for the following six centuries. In the same vein, Putnam (1993) goes far back in history to find the roots of the North–South divide in the persistently low level of social capital (trust) that seems to characterize the Mezzogiorno.

A problem with attributing the relative weakness of Southern industrial growth to historical and geographical factors dating back several centuries is its inconsistency with the initially (1861) relatively small gap in GDP per person. Why did geographical, entrepreneurial, and social capital disadvantages not result in a wider income wedge long before unification? One (unfalsifiable) answer could be that those factors only became important in the context of modern economic growth.

The role of policies never ceased being debated ever since the famous polemical exchanges between Nitti and Fortunato at the turn of the twentieth century, with Nitti (1905) blaming the unified state’s policies for the widening gap between North and South, and Fortunato stressing the original weakness of the Southern economy. The ‘colonial conquest’ of the Mezzogiorno by the Northern elites is where some authors lay the blame for the subsequent economic and social ills of the former kingdom of Naples (for a recent, sophisticated, revival of this approach, see De Oliveira and Guerriero, 2014). As already mentioned, tariff policies have taken centre stage in the debate, as have the various ‘regional policies’ aimed at promoting economic and industrial development in the South.

Some authors (see the reviews in Felice, 2013; Federico and Tena, 2014) explain the North–South divide in terms of a Southern lack of natural resources, and a lower level of human and social capital. A new economic geography approach stresses the importance of natural resources in combination with market access (A’Hearn and Venables, 2013; see also Missiaia, 2014).

All of these factors have probably contributed to the regional divide. To investigate their relative impact, we conducted several correlation exercises.18 The general relationship we tested takes the following form:

(p.135) Yit = α‎ + β‎Xit − 10 + Λ‎i + ε‎it

where Yit, our dependent variable, is per person value added in manufacturing in region i and year t, and Λ‎i are regional controls. The period of investigation runs from 1871 to 1961, and Xit − 10 are five possible explanatory variables lagged by ten years: the degree of water availability, which according to many authors (e.g. Fenoaltea, 2006; data from Ciccarelli and Fenoaltea, 2009) had a huge importance in allowing early industrial production (like silk) and in defining its location; the ease of access both to domestic and foreign markets, measured by a Harris (1954) index of Market Potential, which measures the distance-weighted potential demand faced by each region given its geographical location;19 and endowments of human and social capital, two characteristics of the population often seen as major drivers of economic growth.20

All in all, these correlations suggest that geography-related variables are the most relevant in explaining regional divides. Over the entire period (1881–1961), almost two-thirds of the North–South gap in per capita manufacturing could potentially (p.136) be explained21 by differences in the ease of access to foreign markets. Proximity to domestic markets can explain, at most, a quarter of the final divide; human and social capital turn out to be less relevant in explaining the divergent paths in manufacturing.

6.7 Conclusions

In the late nineteenth century Italy was often regarded as the ‘least of the great powers’ (Bosworth, 1979). Sometimes it was demoted to the rank of ‘first of the small powers’. The same ambiguity existed, at least until the 1960s, regarding Italy’s relative economic position. As Federico and Vasta (2010, p. 229) noted, ‘authors find it difficult to nail Italy down in a simple dichotomy between Core and Periphery: Blattman, Hwang and Williamson (2007) list it among the core countries, alongside the United Kingdom, while Williamson (2008) demotes Italy in the European periphery, with Russia and Spain’. In economic history the meanings of ‘core’ and ‘periphery’ vary according to research goals and time frame, as does the meaning of ‘power’ in political science: for much of the period covered in this chapter, Italy, or parts of it, would fall in the ‘periphery’ or ‘core’ according to the chosen variable of interest.

At the time of unification (1861–70), Italy’s industrial sector was underdeveloped not only in absolute but also in relative terms: while Italy’s GDP per person was about half that of the UK, industrial production per person was only 16 per cent of the UK level. Starting from this condition of typically peripheral industrial backwardness, in the following century Italy realized a pretty much ‘normal’ convergence with the ‘core’ in terms of both industrial output and GDP per person. Given its significant demographic size, Italy remained throughout one of the eight to ten largest world economies, and by the 1980s it had become the fifth or sixth largest world manufacturer.

After a slow start, Italy’s industrialization process broadly followed the path taken by the main European countries: the pace of industrial output accelerated during 1896–1914, proved quite resilient to the shock of war, which brought about significant qualitative changes, increased significantly in the 1920s, and slowed in the 1930s while undergoing major structural changes. Italian industry participated fully in the second post-war golden age of the European economy. Throughout the century, Italian industry grew on average only marginally faster than industry in the world as a whole. The ebb and flow of protectionism also followed a general European pattern. Industrial policy was probably as invasive as, but less efficient than, those in France and Germany. State-owned enterprises and ‘industrial districts’ stand out as idiosyncratic Italian institutions fostering industrialization.

The lack of convergence in regional GDP per person is a well-known peculiarity of Italy’s modern economic growth. While a number of studies exist on regional (p.137) GDP, we have produced for the first time consistent estimates of regional industrial and manufacturing output from 1861 to 1961. Industry progressively concentrated in the Northwestern regions, with the Northeast and Centre roughly maintaining their share of national industry. The South consistently lost ground from 1911 onward. A similar pattern holds for the geographical distribution of the ‘modern’ manufacturing sector (chemical, engineering, and metal-making industries). At the same time, Southern industrial output became increasingly dependent on construction, while manufacturing lost its initial relative weight. When we investigate the causes of the uneven geographical distribution of manufacturing, correlation analyses suggest that the degree of exposure to internal and foreign markets has a better chance of explaining differences in regional per capita manufacturing output than variables such as natural resources, and human and social capital, commonly associated with industrialization.

We have found evidence of a weak convergence among Italian regions in industrial (but not manufacturing) output per capita, which stands in contrast to the lack of convergence for GDP per capita.22 The reasons for the different behaviour may not be the same in the case of peripheral countries, on the one hand, and peripheral regions within Italy, on the other. They form a large research agenda for economic historians, given the emphasis traditionally placed on industrialization as the main driver of modern economic growth. Our findings suggest that the composition of industrial output may impact on the growth of non-industrial sectors: Southern industry did converge on the North’s, but its heavy reliance on construction and traditional labour-intensive industries may not have produced the same effects on agriculture and services as did the modern manufacturing located in the other areas.

To conclude, in 1861–71 Italy was definitely a peripheral industrial country, even if—given its population—it produced about 2.5 per cent of the world’s industrial output. Its movement towards the ‘core’ did not start until the 1890s, but it then progressed ‘as expected’ by its initial backwardness. Japan is possibly the country most similar to Italy in the pattern of per capita industrial growth: its much larger population (about twice that of Italy throughout) made it possible to wage ten years of war against industrially weaker neighbours and four years of total war in the Pacific, starting from a relatively low but nonetheless substantial per capita industrial base. When did Italy (and Japan) join the ‘core’ of the industrial world? There are several possible answers to this question, depending on the yardstick employed. It can be argued that, by the eve of the Second World War, several qualitative features of Italian (and Japanese) industry, in particular the share of ‘modern sectors’ in manufacturing output, were similar to those of the core countries, even though the overall output level lagged behind. Alternatively, one may suggest that by the 1930s, the Northwest was a ‘core’ industrial country, the size of the Netherlands, while the South always remained on the edge between core and periphery.

(p.138) References

Bibliography references:

A’Hearn, B. and Venables, A. J. (2013). Regional disparities: internal geography and external trade. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 599–630.

Baffigi, A. (2013). National accounts 1861–2011. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 157–86 (and appendix thereof, 631–713).

Bairoch, P. (1982). International industrialization levels from 1750 to 1980. Journal of European Economic History 11, 269–333.

Barca, F. (1997). Compromesso senza riforme. In Storia del capitalismo italiano dal dopoguerra a oggi (Ed., Barca, F.). Rome: Donzelli, 3–116.

Battilossi, S., Gigliobianco, A., and Marinelli, G. (2013). The allocative efficiency of the Italian banking system, 1936–2011. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 485–515.

Bénétrix, A., O’Rourke, K. H., and Williamson, J. G. (2015). The spread of manufacturing to the poor periphery 1870–2007. Open Economies Review 26, 1–37.

Blattman, C., Hwang, J., and Williamson, J. G. (2007). Winners and losers in the commodity lottery: the impact of terms of trade on growth and volatility in the periphery 1870–1939. Journal of Development Economics 82, 156–79.

Bonelli, F. (1978). Il capitalismo italiano. Linee generali di interpretazione. In Storia d’Italia. Annali 1. Dal feudalesimo al capitalismo (Eds, Ruggero, R. and Vivanti, C.). Turin: Einaudi, 1193–255.

Broadberry, S., Giordano, C., and Zollino, F. (2013). Productivity. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 187–226 (and appendix thereof, 631–713).

Cafagna, L. (1972). The industrial revolution in Italy 1830–1914. In The Fontana Economic History of Europe, Vol. 4 (Ed., Cipolla, C. M.). Glasgow: Collins/Fontana, 279–38.

Campante, F. and Glaeser, E. L. (2007). Yet another tale of two cities: Buenos Aires and Chicago. National Bureau of Economic Research Working Paper No. 15104.

Caracciolo, A. (1969). Crescita e trasformazione della grande industria durante la prima Guerra Mondiale. In Lo sviluppo economico in Italia (Ed., Fuà, G.). Milan: Franco Angeli, 187–240.

Carreras, A. and Felice, E. (2012). When did modernization begin? Italy’s industrial growth reconsidered in light of new value-added series, 1911–1951. Explorations in Economic History 49, 443–60.

Castronovo, V. (1995). Storia economica d’Italia. Dall’Ottocento ai giorni nostri. Turin: Einaudi.

Ciccarelli, C. and Fenoaltea, S. (2009). La Produzione industriale delle Regioni d’Italia, 1861–1913: una Ricostruzione Quantitativa, Vol. I. Rome: Banca d’Italia.

Ciccarelli, C. and Fenoaltea, S. (2014). La Produzione industriale delle Regioni d’Italia, 1861–1913: una Ricostruzione Quantitativa, Vol. II. Rome: Banca d’Italia.

Ciocca, P. (2007). Ricchi per sempre? Una storia economica d’Italia (1796–2005). Turin: Bollati Boringhieri.

Ciocca, P. (General Ed.) (2011–2014). Storia dell’IRI, Vols 1–6. Rome-Bari: Laterza.

Ciocca, P. (2014). L’IRI nell’economia italiana, Vol. 6. Rome-Bari: Laterza.

Confalonieri, A. (1994). Banche miste e grande industria in Italia, 1914–1933, Vol. I. Milan: Banca Commerciale Italiana.

(p.139) Crafts, N. and Magnani, M. (2013). The golden age and the second globalization in Italy. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 69–107.

Daniele, V. and Malanima, P. (2007). Il prodotto delle Regioni e il divario Nord-Sud in Italia (1861–2004). In Rivista di Politica Economica 97(2), 267–316.

Daniele, V. and Malanima, P. (2014). Perché il Sud è rimasto indietro? Il Mezzogiorno fra storia e pubblicistica. In Rivista di storia economica 1, 3–36.

De Oliveira, G. and Guerriero, C. (2014). Extractive States: The Case of Italian Unification. Amsterdam: ACLE, University of Amsterdam.

Di Nino, V., Eichengreen, B., and Sbracia, M. (2013). Real exchange rates, trade and growth. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 351–77.

Federico, G. (1997). An Economic History of the Silk Industry. Cambridge: Cambridge University Press.

Federico, G. (2005). Seta, agricoltura e sviluppo economico in Italia. Rivista di Storia Economica 2, 123–54.

Federico, G. and Tena, A. (1998). Was Italy a protectionist country? European Review of Economic History 2, 73–99.

Federico, G. and Tena, A. (2014). The ripples of the industrial revolution: exports, economic growth, and regional integration in Italy in the early nineteenth century. European Review of Economic History 18, 349–69.

Federico, G. and Vasta, M. (2010). Was industrialization an escape from the commodity lottery? Evidence from Italy, 1861–1939. Explorations in Economic History 47, 228–43.

Federico, G. and Wolf, N. (2013). A long-run perspective on comparative advantage. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 327–50.

Felice, E. (2013). Perché il Sud è rimasto indietro? Bologna: Il Mulino.

Felice, E. and Vecchi, G. (2013). Italy’s growth and decline, 1861–2011. CEIS Research Paper No. 293, University of Tor Vergata.

Fenoaltea, S. (1988). International resource flows and construction movements in the Atlantic economy: the Kuznets cycle in Italy, 1861–1913. Journal of Economic History 48(3), 605–38.

Fenoaltea, S. (2006). L’Economia italiana dall’Unità alla Grande Guerra. Rome-Bari: Laterza.

Fenoaltea, S. (2014). The fruits of disaggregation: the general engineering industry in Italy, 1861–1913. Collegio Carlo Alberto Working Papers No. 358, June.

Galassi, F. and Harrison, M. (2005). Italy at war 1915–1918. In The Economics of World War I (Eds, Broadberry, S. and Harrison, M.). Cambridge: Cambridge University Press, 276–309.

Galasso, G. (2005). Il Mezzogiorno da ‘questione’ a ‘problema aperto’. Manduria: Piero Lacaita.

Gerschenkron, A. (1955). Notes on the rate of industrial growth in Italy, 1881–1913. Journal of Economic History 15, 360–75.

Gerschenkron, A. (1962). Economic Backwardness in Historical Perspective: A Book of Essays. Cambridge, MA: Belknap Press of Harvard University Press.

Giordano, C. and Giugliano, F. (2012). A tale of two fascisms: labour productivity growth and competition policy in Italy, 1911–1951. Bank of Italy Economic History Working Papers No. 28.

(p.140) Giugliano, F. (2011). Crisis. Which crisis? New estimates of industrial value added in Italy during the Great Depression (mimeo).

Golinelli, R. and Monterastelli, M. (1990). Un metodo per la ricostruzione di serie storiche compatibili con la nuova contabilità nazionale (1951–89). Nota di lavoro n. 9001. Bologna: Prometeia.

Gomellini, M. (2004). Il commercio estero dell’Italia negli anni Sessanta: specializzazione internazionale e tecnologia. Economic History Working Paper, Banca d’Italia, Rome.

Gomellini, M. and Pianta, M. (2007). Commercio con l’estero e tecnologia negli anni Cinquanta e Sessanta. In Innovazione tecnologica e sviluppo industriale nel secondo dopoguerra (Eds, Antonelli, C., Barbiellini Amidei, F., Giannetti, R., Gomellini, M., Pastorelli, S., and Pianta, M.). Rome-Bari: Laterza, 358–594.

Gomellini, M. and Toniolo, G. (2015). The industrialization of Italy in the long-run (mimeo).

Iuzzolino, G., Pellegrini, G., and Viesti, G. (2013). Regional convergence. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 571–98.

James, H. and O’Rourke, K. H. (2013). Italy and the first age of globalization, 1861–1940. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 37–68.

Lafay, G. (1992). The measurement of revealed comparative advantage. In International Trade Modeling (Eds., Dagenais, M. G. and Muet, P. A.). London: Chapman & Hall, 209–34.

Maddison, A. (2010). Statistics on World Population, GDP and Per Capita GDP, 1–2008 AD, online database: http://www.ggdc.net/MADDISON/oriindex.htm

Magnani, M. (1997). Alla ricerca di regole nelle relazioni industriali: breve storia di due fallimenti. In Storia del capitalismo italiano dal dopoguerra a oggi (Ed., Barca, F.). Rome: Donzelli, 500–43.

Malanima, P. (2007). Wages, productivity and working time in Italy (1270–1913). Journal of European Economic History 36 (1), 127–71.

Missiaia, A. (2014). Market access and regional divergence in Italy (1871–1911) (mimeo).

Mitchell, B. R. (2007). International Historical Statistics. Basingstoke: Palgrave Macmillan.

Nitti, F. S. (1905). La ricchezza dell’Italia. Turin: Roux & Viarengo.

OECD (2013). OECD Factbook 2013: Economic, Environmental and Social Statistics. Paris: OECD Publishing.

Putnam, R. D. (1993). Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.

Rodrik, D. (2013). Unconditional convergence in manufacturing. Quarterly Journal of Economics 128, 165–204.

Romeo, R. (1959). Risorgimento e capitalismo. Rome-Bari: Laterza.

Romeo, R. (1988). Breve storia della grande industria in Italia, 1861–1961. Milan: Il Saggiatore.

Salsano, F. and Toniolo, G. (2010). Da Quota 90 allo SME. Rome-Bari: Laterza.

Toniolo, G. (1977). Effective protection and industrial growth: the case of Italian engineering. Journal of European Economic History 6, 659–73.

Toniolo, G. (1980). L’Economia dell’Italia fascista. Rome-Bari: Laterza.

Toniolo, G. (1990). An Economic History of Liberal Italy. London: Routledge.

Toniolo, G. (2013). An overview of Italy’s economic growth. In The Oxford Handbook of the Italian Economy since Unification (Ed., Toniolo, G.). New York: Oxford University Press, 3–36.

(p.141) Toniolo, G., Conte, L., and Vecchi, G. (2003). Monetary union, institutions and financial market integration: Italy, 1862–1905. Explorations in Economic History 40(4), 443–61.

Vitali, O. (1970). Aspetti dello sviluppo economico italiano alla luce della ricostruzione della popolazione attiva. Rome: Tipolitografia Failli.

Williamson, J. G. (2008). Globalization and the Great Divergence: terms of trade booms, volatility and the poor periphery. European Review of Economic History 12, 355–92.

Williamson, J. G. (2011). Trade and Poverty: When the Third World Fell Behind. Cambridge, MA: MIT Press.

Zamagni, V. (1993). The Economic History of Italy, 1861–1990. Oxford: Oxford University Press.

Notes:

(1) Corresponding author: matteo.gomellini@bancaditalia.it. The views expressed herein are those of the authors and do not reflect the views of the institutions represented.

(2) Industry is here defined as including mining and utilities, but excluding construction. It includes cottage industry and the industrial output of small workshops.

(3) For a similar exercise, see Chapter 2 in this volume, and Bénétrix, O’Rourke, and Williamson (2015).

(4) For 1861–1913, Fenoaltea has provided a much more detailed breakdown of industrial value added, which cannot be extended to the rest of our period (Ciccarelli and Fenoaltea, 2009).

(5) For instance, in the nineteenth century and beyond blacksmiths were included in engineering value added, and small blast furnaces in that of metallurgy (Fenoaltea, 2014).

(6) Table 6.1 is from Baffigi (2013), the most recent and reliable available source. It does not, however, provide sectoral breakdowns which, in Table 6.3, are drawn from different sources (hence the discrepancies between the aggregate growth rates in the two tables).

(7) Among recent contributions see Salsano and Toniolo (2010).

(8) In 1884 a steel-making company was created in Terni using electrically powered Bessemer converters and Martin–Siemens furnaces. A typical example of the industrial policy of the left, Terni was created by a group of private entrepreneurs, enjoying the state guarantee of substantial procurement, particularly for the navy. A government committee chose the location, taking into account, among other things, its distance from the border and from the sea (given the maximum reach of naval guns at the time). The largest banks provided most of the needed financial resources.

(9) See among others Guy De Maupassant’s masterpiece, Bel Ami (1885).

(10) Long underrated, Federico De Roberto (1861–1927) is one of the great Italian novelists. His I Vicerè (1894) is a scathing critique of business corruption, as is his posthumous L’Imperio (1929), set at the same time.

(11) On IRI see the six volumes edited by Pierluigi Ciocca (Ciocca, 2011–14).

(12) In 1963, at the peak of the ‘economic miracle’, a macroeconomic ‘democratic planning’ was designed ‘to tackle the main structural problems of the economy […] On the whole the program was heavily influenced by a dirigiste approach not uncommon in European economic culture of the time’ (Crafts and Magnani, 2013, p. 81). Incomes policy proposals ran up against the weakness of the reform culture of the Communist Party (Magnani, 1997); the targets set by the programmazione were hardly compatible with the inefficiency of public administration; and the antitrust regulation and company-law reform proposals were defeated by the consolidated interests of large industrial groups (Barca, 1997; Ciocca, 2007).

(13) Among some of the most recent contributions are Iuzzolino, Pellegrini, and Viesti (2013), A’Hearn and Venables (2013), Felice (2013), and Daniele and Malanima (2007; 2014).

(14) Data are at 1911 constant prices, which may affect the shares of manufacturing and construction presented below.

(15) This is the methodology Fenoaltea used in his ‘first generation’ estimates (see Ciccarelli and Fenoaltea, 2009). There are, obviously, two major flaws in this procedure: it does not take into account unknown but potentially large productivity differentials between regions, and it assumes the same ratio of labour force to employment across regions.

(16) For 1971 we rely on the Centre for North South Economic Research (CRENoS) dataset: www.crenos.it.

(17) See Toniolo (2013) and the contributions therein.

(18) Correlations are obtained by regressing final-year levels (or the decadal growth) of per capita manufacturing on initial-year levels of each explanatory variable. These correlations are robust to a series of controls.

(19) Proximity to foreign markets reduces transportation and information costs, and facilitates imitation and the transfer of technology.

(20) A more accurate description of the variables and details on the econometrics is in Gomellini and Toniolo (2015).

(21) The potential explanatory power can be interpreted as the highest share of the North–South differential in the dependent variable that can be explained by a single variable. It is calculated as in Campante and Glaeser (2007).

(22) On the distinction between convergence in GDP per capita and manufacturing output per capita, see Rodrik (2013).