Introduction

Between 1270 and 1870 Britain slowly progressed from the periphery of the European economy to centre-stage of an integrated world economy. In the process it escaped from Malthusian constraints and by the eighteenth century had successfully reconciled rising population with rising living standards. This final chapter reflects upon this protracted but profound economic transformation from the perspective of the national income estimates assembled in Part I and analysed in Part II of this book. Because Britain’s economic rise did not unfold in isolation, account is taken of the broader comparative context provided by the national income reconstructions now available for several other Eurasian countries: Spain from 1282, Italy from 1310 and Holland from 1348, plus Japan from 725, China from 980 and India from 1600. All are output-based estimates but have been derived via a range of alternative approaches according to the nature of the available historical evidence. Several make ingenious use of real wage rates and urbanisation ratios (Malanima, 2011; Álvarez-Nogal and Prados de la Escosura, 2013), two economic indicators often used as surrogates for estimates of GDP per head. Only the GDP estimates for Holland, like these for Britain, have been made the hard way, by summing the weighted value-added outputs of the agricultural, industrial and service sectors and then dividing the results by estimates of total population obtained by reconciling time-series and cross-sectional demographic data. Methodologically, the British and Dutch national income estimates are therefore the most directly comparable. Each is free from overdependence upon any single or narrow range of data series and, instead, they encapsulate variations in the wide range of economic indicators, appropriately weighted in line with their importance in overall economic activity, from which they have been reconstructed.

Introduction

Agriculture was for long the single largest component of the English and British economies, both in terms of its share of employment and the value of its output. The latter was a function of the amount of land under cultivation, the uses to which it was put, the productivities of crops and animals and their respective prices. The main purpose of this chapter is to describe the methods used to derive the areas under arable and grass and, in particular, the total sown acreage. The crops produced and animals stocked are the subjects of the following chapter. Along the way, it will be demonstrated that claims that the peak arable area in the medieval period may have exceeded 20 million acres (Clark, 2007a: 124) are unrealistic, since, on the best available evidence, the combined total under field crops and fallow could not have been more than 12.75 million acres. In the absence of significant food imports, this limited both the population that could be supported and the supply of kilocalories per head needed for survival. It also shaped the production choices made by agricultural producers.

Comprehensive national agricultural statistics were collected annually from 1866 and provide the starting point for calculating the acreages of arable and grass (Anon, 1968; Coppock, 1984). Together with the tithe files, which provide a precise but incomplete guide to the share of land in each county devoted to arable production during the 1830s (Kain, 1986; Overton, 1986), they are used to provide a nineteenth-century benchmark. The chapter proceeds as follows. After a discussion of the potential agricultural area of England in Section 2.2, Section 2.3 reviews the arable acreage by county from the tithe files of the 1830s and from the agricultural statistics of 1871. Section 2.4 then examines changes in land use between 1290 and 1871, while Section 2.5 presents county-level estimates of the arable acreage in 1290. Section 2.6 provides a further cross-check by examining changes in land use between 1086 and 1290. Finally, Section 2.7 provides estimates of land use for a number of benchmark years between 1270 and 1871.

Introduction

Economic growth can be either extensive or intensive. Extensive growth arises where more output is produced in line with a growing population but living standards remain constant, while intensive growth arises where more output is produced by each person. In the former case, there is no economic development, as the economy simply reproduces itself on a larger scale: in the latter, living standards rise as the economy goes through a process of economic development. To understand the long-run growth of the British economy reaching back to the thirteenth century therefore requires knowledge of the trajectories followed by both population and GDP. Of particular interest is whether periods of intensive growth, distinguished by rising GDP per head, were accompanied by expanding or contracting population. For it is one thing for living standards to rise during a period of population decline, such as that induced by the recurrent plagues of the second half of the fourteenth century, when survivors found themselves able to add the land and capital of those who had perished to their own stocks, but quite another for living standards and population to rise together, particularly given the emphasis of Malthus [1798] on diminishing returns. Indeed, Kuznets (1966: 34–85) identified simultaneous growth of population and income per head (i.e. the concurrence of intensive and extensive growth) as one of the key features that distinguished modern from pre-industrial economic growth.

Introduction

Chapter 6 has argued that workers responded to changes in real wage rates by adapting how hard they worked so as to maintain their earnings. Household incomes therefore tracked GDP per head rather than real wage rates and progressively improved over time, doubling between the early fourteenth and late seventeenth centuries and doubling again over the course of the industrial revolution. Higher incomes translated into changing patterns of consumption and the forms these consumption choices took are the subjects of this chapter. Section 7.2 reconstructs the kilocalorie value and composition of diets based on the agricultural-output estimates presented in Chapter 3, augmented by information on imported foodstuffs. Given that populations require an average daily food intake per head of 2,000 kilocalories (Livi-Bacci, 1991: 27) to provide sufficient nourishment for both economic and biological reproduction, these calculations also provide a useful cross-check on the consistency of the agricultural-output and population estimates. Section 7.3 then considers non-food consumption drawing upon early modern evidence of material culture as revealed by probate inventories. Again, these trends need to be consistent with those of industrial output reconstructed in Chapter 4.

Price, habit, fashion and status all shaped the budgetary decisions taken by households. Demand for food was inelastic up to the point where basic subsistence needs had been met, but as incomes rose there were clear trade-offs to be obtained between increasing consumption of cheap sources of kilocalories such as pottage, potatoes and salted herrings on the one hand, or indulging in more expensive refined bread, quality ale and beer, dairy produce and meat, plus the imported luxuries of wine, sugar, tea, cocoa and tobacco, on the other. In effect, higher incomes allowed more households to trade up to a respectability basket of foodstuffs providing a more varied and processed diet but not necessarily more kilocalories. The changing relative prices of arable, livestock and luxury products influenced these consumption decisions, while the relative cheapness or dearness of food determined how much disposable income could be devoted to the increasingly varied and tempting array of non-food consumer goods (Figure 5.02).

Introduction

Income distribution in England between 1270 and 1870, as elsewhere in Western Europe, was profoundly unequal due to entrenched inequalities in access to the land, capital, education and political power upon which personal wealth depended. Gender, rank and servility and their differential legal rights were determined at birth. Privilege, patronage and position ensured that rent-seeking was rife, while warfare created opportunities for ransom and plunder to the enrichment of those in command and impoverishment of the vanquished. Everywhere, as a result, there were rich men in their castles and poor men at their gates. Moreover, as van Zanden (1995) and Milanovic and others (2007) have demonstrated, the effect of economic growth was to magnify rather than mitigate these inequalities and widen the income gap between those at the top and bottom of the social pyramid.

The rich became richer as average wealth grew because the more wealth there was the greater the opportunities for those with power and privilege to enrich themselves at the expense of the weak and disadvantaged majority. In Holland one legacy of the prosperity achieved during the Dutch Golden Age was a greatly increased inequality of incomes, which was more marked in towns than rural villages and greatest of all in major cities (van Zanden, 1995). In England, similarly, Milanovic and others (2007) claim that inequality rose with average incomes between 1688 and 1801/03, thereby confirming Kuznets’ (1955) observation that income inequality typically increased during the early stages of economic growth and only declined relatively late in the modernisation process. Prior to 1870, therefore, increasing inequality can be treated, like urbanisation, as a characteristic and unavoidable manifestation of economic growth.

Introduction

This chapter provides annual estimates of output in agriculture, which was the largest sector of the economy during the middle ages, and continued to play an important role throughout the period under consideration. The approach builds on the study of Overton and Campbell (1996), which tracked long-run trends in agricultural output and labour productivity, but was restricted to estimates for a small number of benchmark years. To provide annual estimates, heavy reliance has been made on three datasets assembled for the late-medieval, early modern and modern periods. For the period c.1250 to c.1500, a Medieval Accounts Database has been assembled by Campbell (2000, 2007), drawing upon the archival labours of a number of other historians, including David Farmer, John Langdon and Jan Titow. The information on arable yields and animal stocking densities is taken largely from manorial accounts, but is supplemented by information on the non-manorial sector from tithes. For the period c.1550 to c.1750, an Early Modern Probate Inventories Database has been assembled by Overton, which provides animal stocking densities and indirect estimates of arable yields from the valuation of the assets left by farmers (Overton and others, 2004). From the early eighteenth century, use is made of the Modern Farm Accounts Database assembled by Turner, Beckett and Afton (2001).

The chapter proceeds as follows. Section 3.2 provides a brief introduction to the main data sources for the three periods. Estimates of output for the arable sector are then given in Section 3.3, followed by estimates of livestock-sector output in Section 3.4. The arable and livestock outputs are combined in Section 3.5 to provide estimates of overall agricultural output, while Section 3.6 concludes.

Introduction

In 1270 the agricultural sector dominated economic output, dwarfing the industrial and service sectors. By 1870, notwithstanding an eightfold expansion of agricultural output, this situation had been reversed and industry and services were the fastest-growing and largest sectors. The progress of British industry has been closely scrutinised from 1700 but less so in earlier centuries notwithstanding that the roots of Britain’s industrial rise extend back much earlier than the conventional starting date of the industrial revolution in the mid-eighteenth century. The service sector, which already by the mid-nineteenth century had overtaken industry and emerged as the dominant sector within the economy, has received far less attention and awaits systematic investigation from the bottom up. This unevenness of treatment has required adoption of a range of approaches in order to derive valid estimates of industrial and service-sector output and thereby chart these profound changes in the structure of economic activity and volumes of industrial and service-sector output across the 600 years under investigation.

Industrial output

From 1700 industry is the one economic sector for which annual data have previously been gathered and analysed on a national scale. Full use has therefore been made of these existing estimates. Pioneering work by Hoffmann (1955) inadvertently overstated the growth rate of industrial output during the industrial revolution as a result of the weighting procedures applied to a dataset which covered only 56 per cent of industrial output. As Harley (1982) and Crafts (1985) separately point out, the problem is that a few industries, most notably cotton and iron, grew more rapidly than the rest of manufacturing, and these atypical industries bulk disproportionately large in Hoffmann’s output series. By extrapolating total industrial output from that series he effectively doubled the weights of the most dynamic industries. Harley (1982) and Crafts and others (1989) have overcome this problem by limiting the weights applied to cotton and iron and increasing those applied to other industries, thereby arriving at lower estimates of total industrial output growth.

Introduction

How does Britain’s experience of long-run economic growth and development, as revealed by the output-based estimation of GDP per head set out in Part I of this book, compare with that of other countries? Maddison’s (2010) historical national income estimates show that by the middle of the nineteenth century Britain had become the most developed economy in the world, with higher output per head than any other country in Europe, Asia or the Americas. A majority of its population lived in towns, agriculture contributed less than a quarter of employment and a fifth of value-added output, after centuries of mercantilism it was trading across the world under the banner of free trade, and the value of that international commerce accounted for a fifth of national income and was rising. Demographic and economic growth were proceeding in tandem and thereby fulfilling one of Kuznets’s (1966) key requirements of modern economic growth. Contrary to Malthus’s gloomiest predictions, the population was not only growing but it was becoming richer. The Great Exhibition of 1851, conceived to make clear to the world Great Britain’s role as industrial leader, could not have been better timed.

Eight centuries earlier, when William of Normandy had cast his covetous eyes upon the Crown of England, the country had been less a land of plenty than a kingdom with plenty of land. Its relatively sparse population of 1.7 million was overwhelmingly rural, towns were small and London alone had more than 10,000 inhabitants, commerce was limited and commercial institutions and infrastructure weakly developed, and exports were chiefly of unprocessed primary products, most notably wool and tin. England may have been resource-rich but its lack of development meant that its GDP per head was only a quarter what it would become in 1850. It was poorer than most of its immediate continental neighbours, significantly poorer than northern and central Italy, at that time Europe’s economic leader, conspicuously poorer than the world’s most successful economy, China under the Northern Song Dynasty (960–1127), and poorer than the core economies of the Roman Empire a millennium earlier under Augustus (Lo Cascio and Malanima, 2009).

Introduction

England’s aggregate output per head doubled between the 1290s and the 1690s, trebled by the 1800s and had quadrupled by the 1850s: self-evidently, the productivity of labour was rising across this long period. Whether it was rising equally across all three sectors is another matter. Deane and Cole’s (1967) once influential account of British economic development since 1688 was premised on an eighteenth-century agricultural revolution releasing labour to industry at the very time that mechanisation and the division of labour were raising the productivity of labour in manufacturing. British-made goods thereby became unbeatable in world markets so that industry became the most dynamic employment sector within a fast-growing economy. Yet evidence of agricultural innovation a century earlier (Kerridge, 1967) suggested that agricultural output and probably labour productivity were rising from the seventeenth century, so that the share of the labour-force in agriculture was already much reduced before the industrial revolution got under way. Initially controversial, this view has since gained considerable legitimacy (Allen, 1992, 1999; Overton, 1996, 80–2, 121–30). Indeed, it would be difficult to account for an English urbanisation ratio which rose from 5.25 per cent in the 1520s to 17.0 per cent by 1700 and then 27.5 per cent in 1801 (Wrigley, 1985: 688) if there were no concurrent improvements in agricultural labour productivity.

The post-1700 output estimates of Crafts (1985) and Crafts and Harley (1992) accepted the revisions of the agricultural historians and, in turn, proposed that industrial growth was slower during the eighteenth century than estimated by Deane and Cole, notwithstanding the continued transfer of labour out of agriculture and into industry. On this scenario eighteenth-century agriculture was more successful at shedding labour than industry was at expanding output: hence the paradox that, at the very time that Britain was becoming the workshop rather than the granary of the world (Crafts, 1989), productivity growth in agriculture apparently exceeded that in industry.

Gross Domestic Product (GDP) is the single most widely employed measure of the value of a country’s market-based economic activity, as GDP per head is of relative and absolute levels of prosperity, and annual rates of change in GDP per head are of the pace of economic growth. GDP has its flaws (it omits non-market activity and leisure, and captures changes and differences in quality, especially of services, imperfectly) but has the merit of being widely understood and respected (Leunig, 2011: 358). There is no alternative single measure that does the same job more effectively. That is why today many governments make their own estimates of GDP per head based upon official statistics of economic output and population and why the United Nations, World Bank and other organisations publish annual estimates ofGDP andGDP per head for most of the world’s economies, including many for which only the most rudimentary economic and demographic data are available. It is these estimates that inform contemporary debates about the pace of economic growth, widening gap between rich and poor countries, and progression of countries from underdevelopment to development. Obtaining a proper historical perspective on these issues is more problematic, for governments took little interest in the gathering of official statistics before the nineteenth century and the first attempts to measure GDP followed some time later. It has therefore devolved upon historians to rectify this deficiency, drawing upon a range of mostly unofficial data sources.

Publication in 1962 of Phyllis Deane and Max Cole’s British economic growth, 1688–1959 marked a watershed in historical analysis of economic growth. Simon Kuznets and his colleagues at the National Bureau of Economic Research had already applied the relatively new techniques of national income accounting to the measurement of economic growth since the late nineteenth century, when the modern statistical age effectively began. Deane and Cole’s innovation was to extend national accounting methods to investigation of a period spanning 272 years and beginning long before statistical agencies produced long time-series of data on consistently defined variables. To achieve this they assembled their own datasets, made ingenious use of proxy measures when direct evidence was lacking, and modelled what was missing altogether. Their book is a mine of information and a model of clarity and logic.

It is fair to say that this new approach to economic history before the mid-nineteenth century was not welcomed by all economic historians, and some reviewers focused their attention on the shortcomings of the data series constructed by Deane and Cole, which they saw as undermining the credibility of the conclusions being drawn about the processes of British economic development. Other economic historians, inspired by Deane and Cole’s novel approach but sceptical of the authors’ findings, responded by refining and extending the available datasets, seeking better ways of combining them into robust estimates of national income, and ensuring that all assumptions made were empirically well grounded. After 50 years of work in this vein, Deane and Cole’s basic analysis has been fairly comprehensively revised and understanding of the processes of economic growth during the world’s first industrial revolution has been elevated to a new plane.

Introduction

Part I of this book has presented a predominantly positive picture of long-run economic growth and development in Britain from the Black Death of 1348–9 until 1870. Between the early fourteenth century and 1700 GDP per head approximately doubled and it doubled again between 1700 and 1870. Before 1780 progress was fitful, with the greatest gains concentrated into the twin periods of demographic decline in the second halves of the fourteenth and seventeenth centuries, but, crucially, there was little erosion of these gains during the sequel episodes of population growth in the sixteenth and eighteenth centuries. In fact, well before the industrial revolution got under way population and GDP per head were rising together. Malthus [1798], however, in his Essay on the principle of population, was convinced that the relationship between population growth and output per head was otherwise, since, sooner or later, diminishing returns to labour were bound to accrue. This is possibly exemplified by the inverse correlation between population and GDP per head that appears to have prevailed during the second half of the thirteenth century. It is without parallel over the next five centuries: later phases of population growth certainly brought their quota of socio-economic difficulties but, for the most part, declining GDP per head was not one of them.