EH 1 SB TOPIC 1 Economic Growth 1000

EH 1: SB TOPIC 1 Economic Growth, 1000 -1870

OVERVIEW • This part of the course aims to provide an overview of the Great Divergence of productivity and living standards between Europe and Asia • Emphasis is placed on the need to establish firmly the historical “facts” that are to be explained as well as on providing a satisfactory economic explanation 2

TOPICS • • 1. Economic Growth, 1000 -1870 2. Wages, Prices and the Great Divergence 3. Cotton Textiles and the Great Divergence 4. Accounting for the Great Divergence 3

TOPIC 1: ECONOMIC GROWTH, 1000 -1870 LECTURE OUTLINE A. HISTORICAL NATIONAL ACCOUNTING 1. The Maddison Data Set 2. Basis of the Population Estimates 3. Basis of the Income Estimates 4. International Comparisons 5. International Comparisons before 1870: The Great Divergence Debate • 6. Improving on Maddison • • • 4

A. HISTORICAL NATIONAL ACCOUNTING • Until surprisingly recently, there was no systematic quantitative overview of the basic contours of long run economic development • Although there were a few isolated studies with data on individual parts of a number of European economies, there had been no attempt to pull them together • The pioneer here was Angus Maddison, who spent a lifetime developing an approach known as “Historical National Accounting” 5

Historical National Accounting • I will end up suggesting that Maddison’s estimates are seriously misleading and are in the process of being replaced • But that should not take away from the magnitude of his achievement. He was unambiguously the pioneer of a great project, which remains to be completed 6

1. THE MADDISON DATA SET • Quantification of world economic history received a major boost with the publication of Maddison’s (2001), World Economy: A Millennial Perspective • Until the early 1990 s, data on comparative levels of per capita income were largely restricted to the “Maddison 16” of rich industrialised countries and covering only the period since 1870 7

The Maddison data set • Following an explosion of interest in the production of historical national accounts, by 1995 Maddison was able to extend estimates back to 1820 for many rich industrialised countries and back to 1870 for many other countries • Maddison’s 2001 volume is much more ambitious, and aims to provide global coverage for the period back to 1820 and rough estimates back to 1000 for selected economies • Final version from 2010 available on GGDC website 8

What Maddison’s data show • TABLES 1 & 2: Between 1000 and 1998, population rose 22 -fold, per capita income rose 13 -fold. In previous millennium, population increased by onesixth and per capita income fell slightly. • Two distinct epochs in 2 nd Millennium: – Between 1000 and 1820, p. c. income increased by 50%, population rose by a factor of four. Growth was largely extensive – Since 1820, more intensive development (g of p. c. income > g of pop). Pop increased by factor of < 6, p. c. income increased by factor of > 8 9

TABLE 1: Level and rate of growth of population: World and major regions, 0 -1998 A. D. 10

TABLE 2: Level and rate of growth of GDP per capita: World and major regions, 0 -1998 A. D. 11

What Maddison’s data show • Wide disparity in performance across regions: – Group A: W. Europe, W. Offshoots, Japan – Group B: Latin America, E. Europe & Former USSR, Asia (except Japan), Africa • Growth of p. c. income 4 times as fast in Group A as in Group B. – – Group B was ahead in 1000 (after fall of Roman Empire). By 1820 p. c. income in Group A twice level in Group B. By 1998 ratio had increased to 7: 1 In 1998, p. c. income ratio between W. Offshoots and Africa more than 19: 1. Africa today at W. European level of 1820 12

What Maddison’s data show • For Group A countries, at least, Malthus was clearly too pessimistic about the possibility of increasing living standards in the long run • TABLE 3: Changing regional shares of GDP: – In 1000, Asia produced two-thirds of world GDP and W. Europe less than 9% – By 1820, W. Europe’s share 24%, Asia’s share 56% – By 1998, Asia’s share 30%, W. Europe and W. Offshoots 46% 13

TABLE 3: Level and rate of growth of GDP: World and major regions, 0 -1998 A. D. 14

Europe’s Little Divergence, 15001820 • Maddison’s data break down developments within Western Europe from 1500. Pattern is sometimes described as the Little Divergence • Italy had highest per capita income during 1000 -1500 • After “voyages of discovery”, brief Iberian boom, but opening up of Atlantic economy led eventually to economic leadership of NW Europe: first NL, then GB 15

TABLE 4: GDP per capita in UK and selected countries, 1500 -1820 (1990 international dollars) 16

2. BASIS OF THE POPULATION ESTIMATES • Data from the past exist largely as a result of government efforts to raise taxes • China: bureaucracy kept population registers going back more than 2000 years. Designed to assess taxable capacity, also included information on cultivated area and crop production • Roman Empire: conducted censuses for fiscal purposes 17

Population • England: Domesday Book 1086 extraordinarily detailed survey of whole country, broken down by estate. • A number of economic historians have used Domesday data to estimate English population in 1086 (Russell, 1948; Darby, 1977) • Others have also used Domesday Book to estimate national income for England in 1086 (Snooks, 1995; Mayhew, 1995; Walker, 2015) 18

Population • Parish registers in many countries record births, marriages & deaths • Decadal censuses now the norm in many countries • Scope for disagreement small after 1700 • Even between 1000 and 1700, fair amount of agreement between scholars • There is some scope for improving on Maddison’s population estimates, but it is probably more about increasing the frequency of observations than changing the trends 19

3. BASIS OF THE INCOME ESTIMATES • For many countries, historical national accounts have been reconstructed since about 1870 • Going back further in time, for “core” countries there is quite a lot of hard evidence that can be derived from archives: – cultivated land use, crop yields and animal numbers from farm accounts and tithe receipts – population living in towns and labour force occupations – extent of foreign trade – wages, prices and real wage trends – government spending and revenue 20

Basis of the income estimates • However, despite encouraging economic historians to produce GDP estimates based on hard data, Maddison preferred to rely largely on “guesstimates” for the pre-1870 period • His approach was to use some basic assumptions, particularly about the subsistence level of income, and make some freehand adjustments on the basis of the qualitative literature • This is best understood if we consider how Maddison made his international comparisons and established the subsistence level of income 21

4. INTERNATIONAL COMPARISONS • It is important to compare wages and GDP per capita at Purchasing Power Parity (PPP) • British wages will be in £, while US wages will be in $ • Cannot simply compare at the exchange rate, which may move to clear international asset markets, rather than reflecting the price of goods and services that people living in the 2 countries buy 22

International comparisons • Compute the PPP by comparing prices of individual products, weighted by their importance in the GDP of the 2 countries • Would get a different PPP if use UK weights rather than US weights • A compromise estimate is obtained by using the geometric mean of UK and US weights • However, a series of bilateral comparisons made this way may not be transitive 23

International comparisons • Maddison therefore chose to use a set of international weights for comparing all countries • Most international comparisons are now carried out in 1990 GK international dollars (named after Geary and Khamis who derived the weighting scheme) 24

1990 international dollars • Using PPPs for 1990, we can convert GDP per capita for all countries to 1990 international dollars • This provides a convenient standard for comparing per capita incomes over space and time • World Bank standard in 1990 suggested “bare bones subsistence” is $1 per day, or $365 per year • Since any society has a rich elite, minimum average per capita income for any society likely to be around $400 • This is widely regarded as “bare bones subsistence” 25

1990 international dollars • Now look at the early observations in the Maddison data set and notice how many are at or very close to $400 in the early years • Better off economies are put slightly above this level, at $425 or $450 • Observations between 1000 and 1820 are then more or less drawn in freehand, so as to conform to Maddison’s views on the major controversies in global economic history 26

5. INTERNATIONAL COMPARISONS BEFORE 1870: THE GREAT DIVERGENCE DEBATE • A central issue in international comparisons before 1870 concerns the Great Divergence of living standards between Europe and Asia • But which Asian regions should be compared with which European countries? • Maddison compares W. Europe with China, but the Yangzi Delta region of China may not have been so far behind even by 1800 27

Maddison’s Great Divergence • Pomeranz claims the Yangzi delta was on a par with GB & NL in 1800, but this has not been widely accepted • Maddison’s figures for W. Europe and China show a more conventional picture (TABLE 5): – After the fall of the Roman Empire, W. European p. c. income fell to a low point circa 1000. – Chinese incomes were stagnant, so W. Europe fell behind 28

TABLE 5: Comparative levels of GDP per capita in China and Western Europe, 400 -1998 A. D. (1990 international dollars) 29

Maddison’s Great Divergence – P. c. income in W. Europe began to grow from C 11 th, led by N. Italian city states (esp. Venice), then Portugal & Spain, then NL, then GB – A growth rate of 0. 14% p. a. was not very fast but by 1820 p. c. income in W. Europe had tripled (compound interest) – Chinese p. c income also grew between 1000 and 1300, but more slowly than W. Europe. So W. Europe caught up with China by C 14 th 30

Maddison’s Great Divergence – Chinese p. c. income then stagnated until 1820. Initially due to internal policies and institutions (e. g. self-imposed isolation), then reinforced by colonial exploitation by West – China fell far behind as W. Europe continued to grow. By 1820, p. c. income in W. Europe twice Chinese level – From early 19 th to mid-20 th century, Chinese p. c. income fell, while W. European growth accelerated 31

Maddison’s Great Divergence – By 1950, W. Europe’s p. c. income more than 10 times Chinese level – Since 1950, and esp since 1973, China has been catching up. By 1998, W. European income level less than 6 times Chinese level 32

6. IMPROVING ON MADDISON • Many of Maddison’s GDP per capita observations before 1820 are “guesstimates”, but it doesn’t have to be this way • Since Maddison died in 2010, the Maddison Project has begun the process of improving upon the original Maddison data set • Bolt and van Zanden (2014) represents the first round of improvement, and incorporates a number of series for the pre-1820 period based on hard data • Figure 2 shows the results for 4 European economies 33

TABLE 6: GDP per capita levels in Europe (1990 international dollars) Sources: Broadberry et al. (2015); van Zanden and van Leeuwen (2012); Malanima (2011); Alvarez-Nogal and Prados de la Escosura (2013) 34

Improving on Maddison • The most obvious change is that levels of per capita income were substantially higher in 1500 than suggested by Maddison • Projecting back further suggests that per capita incomes in Western Europe were already well above bare bones subsistence in the 13 th century, and indeed even earlier – English p. c. GDP in 1086 was about the same as in 1300 35

Improving on Maddison • In the second half of this lecture, we will explore the new British historical national accounts: – Explain where the estimates come from – Explore the implications for understanding the transition to modern economic growth 36

TOPIC 1: ECONOMIC GROWTH, 1000 -1870 LECTURE OUTLINE • • • B. BRITISH ECONOMIC GROWTH 1. Agriculture 2. Industrial Output 3. Services 4. Real GDP 5. Population and GDP per capita 6. Consumption and Output 7. Income and Output-Based Measures 8. Britain in a Comparative Perspective 37

B. BRITISH ECONOMIC GROWTH • Broadberry, Campbell, Klein, Overton & van Leeuwen (2011; 2015): Reconstruction of national income in Britain, 1270 -1870 • GDP estimates combined with population data to derive per capita GDP • Provides a unified framework for addressing the key issues of British economic history before and during the transition to modern economic growth 38

39

Results • Annual growth rate of GDP per capita 0. 17% 12701700 cumulates to doubling of living standards • Annual growth rate of GDP per capita 0. 48% 17001870, cumulates to a further doubling of living standards • If project back from 1850 level, per capita income in late Middle Ages around $1, 100 in 1990 prices, well above “bare bones subsistence” ($400) • High share of livestock agriculture already in late Middle Ages, providing varied diet rather than high level of kilocalories 40

Contrast with Clark • Clark (2010): “A Farewell to Alms argued that England in 1800 was no richer than the average hunter-gatherer society, and no richer than preindustrial England through most of its history since 1200. ” • International comparative framework: if nothing happened between 1200 and 1800, how did Britain go from being an economic backwater to the centre of the global economy? 41

Kalahari bushmen 42

England circa 1200 43

England circa 1800 44

1. AGRICULTURE • Data sources for 3 periods yield information on land use, arable yields, animal stocking densities etc: – Medieval period: Bruce Campbell’s Manorial Accounts Database – Early modern period: Mark Overton’s Probate Inventories Database – Modern period: Turner, Beckett and Afton’s farm accounts database 45

Agricultual output trends • FIGURE 1: Constant price output for arable and livestock sectors • Medieval period: arable output trended down, livestock output stable. Total agricultural output showed modest decline • From mid-C 16 th livestock first grew slower then faster than arable from mid-C 17 th • Total agricultural output growth increased but only modestly 46

FIGURE 1: Indexed output in English arable and livestock agriculture (log scale, 1700=100) 640 320 arable output pastoral output 160 80 40 20 10 1270 1370 1470 1570 1670 1770 1870 47

Share of livestock agriculture • Share of livestock sector rising in constant price terms, but with setback 1450 -1650 • Price of livestock products declined relative to arable products 1450 -1650, so share of livestock sector declined much further in current prices • Note very high share of livestock sector in late medieval period in both current and constant prices • Although kilocalories person not particularly high, expensive kilocalories (meat, dairy produce, ale) 48

TABLE 5: Output shares in English agriculture, in current prices, 10 -year averages (%) 49

FIGURE 2: Indexed output of agriculture, industry and services, England (1270 -1700) and Great Britain (1700 -1870), (log scale, 1700 = 100) 2048 1024 512 256 128 64 agricultural output 32 industrial output 16 services output 8 4 1270 1310 1350 1390 1430 1470 1510 1550 1590 1630 1670 1710 1750 1790 1830 1870 50

2. INDUSTRIAL OUTPUT • ENGLAND, 1270 -1700 • Volume measures available for key industries, combined with value added weights for c. 1700 • Sharp decline across Black Death clearly visible, but stability as population declined further • Return to strong growth after 1500 51

Industrial output in GB, 1700 -1870 • Our industrial production index very close to Crafts/Harley “revised best guess index” • Slower growth than Hoffmann index, which over-weighted cotton • Output growth accelerated from 1760 s to 1830 s before tapering off 52

3. SERVICES • ENGLAND, 1270 -1700 • Deane & Cole approach: – government (state finance data) – commerce (volume indicators) – housing and other services (grew in line with population) • Services trended down with population after Black Death before picking up strongly between 1500 and 1650 53

Services in GB, 1700 -1870 • Deane & Cole approach again • Results broadly consistent with Crafts, but annual frequency • Increase in growth rate from 1780 s 54

4. REAL GDP • Sectoral output weights derived by reflating real output data with sectoral price indices and projecting back in time from 1841 shares • Before 1700, agricultural, industrial and service prices tended to move together over the long run but with some significant shorter swings • After 1700, industrial prices declined sharply relative to other prices 55

FIGURE 3: Aggregate and sectoral price indices, England (12701700), Great Britain (1700 -1870), (log scale, 1700=100) 512 agriculture 256 industry services 128 GDP deflator 64 32 16 8 1270 1320 1370 1420 1470 1520 1570 1620 1670 1720 1770 1820 1870 56

Real GDP • Sectoral weights provided for benchmark years where labour force data also available for cross-checking • Agriculture’s share declined, while services increased the most, particularly after 1700 when relative price of industrial goods fell sharply • But industry grew fastest in constant price terms, both before and after 1700 57

TABLE 9: Sectoral shares in current price GDP, 12701870 (%) 58

5. POPULATION AND GDP PER CAPITA • ENGLISH POPULATION • English population before 1541 controversial • We use data on trends in individual manors or parishes c. 1317 -1541, building on the approach of Hallam for the period 1086 -1317 • Peak medieval population 4. 81 million in 1348, above Overton and Campbell’s estimate of 4. 0 to 4. 5 million but well below Richard Smith’s 6 million • National accounts perspective: difficult to see how many more could have been fed, given food availability 59

English population, 1250 -1700 60

Real GDP per capita in England, 12701700 • Modest positive per capita income growth 1270 -1700, at average annual rate of 0. 17% – 1270 s-1340 s: stationarity of p. c. income – 1340 s-1400 s: strong growth after Black Death – 1400 s-1650 s: plateau – 1650 s-1690 s: return to growth 61

FIGURE 4: Indexed GDP, population, and per capita GDP, England (1270 -1700) and Great Britain (1700 -1870): (log scale, 1700 = 100) 1, 024 512 256 128 64 GDP population 32 16 GDP/cap 1270 1310 1350 1390 1430 1470 1510 1550 1590 1630 1670 1710 1750 1790 1830 1870 62

Real GDP per capita in Great Britain, 1700 -1870 • Per capita income growth averaged 0. 48% between 1700 s and 1860 s, but again episodic: – 1700 s-1830 s: growth slower than during 2 nd half of C 17 th – From 1830 s: sharp acceleration, p. c. income growth comfortably exceeding 1% for first time 63

6. CONSUMPTION AND OUTPUT • Can check level of consumption per head in kilocalories • Livi-Bacci: need 2, 000 kilocalories per day, but some of this from home-raised and wild sources • Slavin estimates 300 kcals from poultry, fish etc. before Black Death, 200 kcals thereafter, which matches with estimates for c. 1900 from Prest • Note that total kcals per head does not trend upwards, but remains constant at around 2, 000. Improvements to diet come through growing share of meat and growth in the extent of processing 64

Per capita daily kilocalorie consumption of major arable crops and animal products in England 65

Consumption • English society under pressure in half century before Black Death. Difficult to see how bigger population could have survived • Further periods of pressure in C 17 th as population recovered and early C 19 th before substantial grain imports • Large livestock sector provided varied diet but not many more kilocalories, so English population does not seem well off if assess living standards in terms of kilocalories • But large proportion of population able to consume meat, dairy produce and ale • In a strongly Malthusian economy, with real wages driven down to bare bones subsistence, bulk of population would derive most of their kilocalories from inferior grains with little processing (e. g. oatmeal) • Social tables suggest 24. 2% unable to afford respectability basket in 1688, 39% in 1290 66

7. INCOME AND OUTPUT-BASED MEASURES • Per capita GDP plotted together with real wages of unskilled building workers in FIGURE 5 • There are differences between the Clark and Allen series, but bigger difference is between both real wage series and GDP per capita • 2 main periods of divergence: “golden age” of C 15 th and “Engels’ pause” of 2 nd half of C 18 th 67

FIGURE 5: Indexed daily real wage of an unskilled building worker and GDP per capita (10 -year moving averages, mean of 1270 -1870=100, log scale) 320 160 80 Allen Clark 40 GDP/cap 20 1270 1320 1370 1420 1470 1520 1570 1620 1670 1720 1770 1820 1870 68

Golden Age of C 15 th and Engels’ Pause • Real wages increased faster than GDP p. c. 1340 s-1450 s then declined between 1450 s and 1650 s while GDP p. c. remained stable • Peak wage rates of mid-C 15 th not seen again before mid-C 19 th, hence important in creating impression of no long run progress • During 2 nd half of C 18 th, real wages flatlined while GDP p. c. continued to grow steadily 69

Real wages and GDP per capita • Angeles (2008) relates changes in real wages to changes in GDP p. c. , labour’s share of income, labour supply person and relative prices: 70

Real wages and GDP per capita • Daily real wage could stagnate while GDP per capita increases if: • (1) Labour’s share of national income falls • (2) People work more days per year • (3) Price of basic consumption goods rises relative to price of all goods 71

Real wages and GDP per capita • TABLE 11: Uses the Angeles decomposition to simulate change in annual days worked, given trends in real wages, GDP p. c. , labour’s share and relative prices • Results for 1750 s-1800 s consistent with those of Angeles: labour supply is most important factor in reconciling positive p. c. GDP growth with falling real wages, although rising relative price of consumer goods also plays a role 72

TABLE 11: Explaining the divergence between GDP per capita and real wages (1700=100) 73

“Golden Age” of C 15 th • 1340 s to 1450 s: real wages able to grow faster than GDP per capita mainly because of labour’s rising share, but also because food was becoming relatively cheap • 1450 s to 1650 s: GDP p. c. stable while real wages fell largely because of “industrious revolution” • Subsidiary roles for increase in relative price of food and fall in labour’s share 74

Variations in Labour Supply • TABLE 12: Independent estimates of days worked per year broadly consistent with this picture of an industrious revolution • Voth uses court records to suggest sharp increase between 1760 and 1800 • Clark and van der Werf use daily and annual wage rates to suggest similar scale of increase, but more gradual process 75

TABLE 12: Estimates of annual days worked person 76

Variations in Labour Supply • Allen and Weisdorf use Blanchard’s data on workers who combined farming and mining in the Mendips to show similar level of days worked in late C 16 th to Clark and van der Werf • Also show much smaller number of days worked in period before post-Reformation reduction in holy days • Growth of annual incomes thus consistent with stagnation in daily real wages because of “industrious revolution” 77

8. BRITAIN IN A COMPARATIVE PERSPECTIVE • Projecting back from Maddison’s 1850 figure of p. c. income in 1990 international dollars suggests for Late Middle Ages a figure of around $1, 100 • This is well above Maddison’s $400 subsistence income, observed for many poorer countries in 20 th century • Little Divergence: Holland then GB forged ahead of Italy and Spain during C 17 th and C 18 th • Asian GDP per capita already much lower in early modern period 78

GDP per capita levels in 1990 international dollars 79

Europe before the Industrial Revolution • Importance of mixed agriculture with a large livestock sector in North Sea area • This meant production had a number of characteristics important for development: – – high value added capital intensive non-human energy intensive enhanced employment opportunities for females, underpinning European marriage pattern • NW Europe pulled ahead as these production techniques spread to industry and services and EMP encouraged human capital formation 80