Homework answers / question archive / Wrigley, E

Wrigley, E

Sociology

Share With

Wrigley, E.A. Urban growth and agricultural change: England and the continent in the early modern period pp. 157-193 Wrigley, E.A., (1987) People, Cities and Wealth: The Transformation of Traditional Society, Oxford: Basil Blackwell Staff and students of London School of Economics and Political Science (LSE) are reminded that copyright subsists in this extract and the work from which it was taken. This Digital Copy has been made under the terms of a CLA licence which allows you to: • access and download a copy; • print out a copy; Please note that this material is for use ONLY by students registered on the course of study as stated in the section below. All other staff and students are only entitled to browse the material and should not download and/or print out a copy. This Digital Copy and any digital or printed copy supplied to or made by you under the terms of this Licence are for use in connection with this Course of Study. You may retain such copies after the end of the course, but strictly for your own personal use. All copies (including electronic copies) shall include this Copyright Notice and shall be destroyed and/or deleted if and when required by London School of Economics and Political Science (LSE). Except as provided for by copyright law, no further copying, storage or distribution (including by email) is permitted without the consent of the copyright holder. The author (which term includes artists and other visual creators) has moral rights in the work and neither staff nor students may cause, or permit, the distortion, mutilation or other modification of the work, or any other derogatory treatment of it, which would be prejudicial to the honour or reputation of the author. Course of Study: EH482 - Pre-Modern Paths of Growth: East and West Compared, c10001800/1900 Title: People, Cities and Wealth: The Transformation of Traditional Society Name of Author: Wrigley, E.A. Name of Publisher: Basil Blackwell 156 Urban growth The comparative neglect of London as a potent engine working towards change in England in the century 1650_175066 is the more paradoxical in that the dominance of Paris within France has long been a familiar notion in political history. Yet London was larger than Paris, was growing much faster and contained a far higher fraction of the national population. All leavens do not, of course, work equally effectively in their lumps; and political dominance connotes different issues from economic and social change, but the irony remains.67 A just appreciation of London's importance must await a fuller knowledge of many points which are still obscure. Meanwhile, this short sketch of a possible model of London's relationship with the rest of the country will have served its purpose if it helps to promote further interest in the complexities of the changes to which no doubt it does only the roughest of justice. 7 Urban Growth and Agricultural Change: England and the Continent in the Early Modern Period The complexity and contingency of any relationship between economic growth and urban growth should need no stressing.1 It is clearly hazardous to undervalue, still more to ignore, the difficulties attending any examination of this topic. In what follows in the early part of this essay I may seem to sail close to the wind in this respect, preferring to sketch out an initial thesis rather starkly at the risk of over-simplifying 'reality'. Later, the complexity of historical experience will ensure that the discussion must be broadened and that the limitations of simple formulations will become evident. Ceteris paribus a rising level of real income and a rising proportion of urban dwellers are likely to be linked phenomena in a pre-industrial economy. If income elasticity of demand for food is less than unity, then with rising real incomes demand for secondary and tertiary products will grow more rapidly than that for primary products and will therefore cause employment in secondary and tertiary industries to rise more rapidly than in agriculture. Such employment is likely to be disproportionately in towns, especially in the case of tertiary employment, and will result in an increase in the proportion of the total population living in towns. There may be an important 'feedback' element in this relationship since the growth of towns may help to further agricultural investment and specialization and so carry forward the rising trend in real incomes.2 Declining real incomes will, of course, tend to have an opposite effect. In an economy which meets its own food needs, urban growth may not only be a symptom of rising real incomes: it may also be a rough measure of the level of productivity per worker in the agricultural sector of the 66 See Laslett, 'The numerical study of English society', pp. 11-12 for a brief discussion of much the same point. 67 It is symptomatic of the neglect of this topic that a work as perceptive and authoritative as Deane and Cole's analysis of British economic growth from the late seventeenth century onwards passes over the growth of London almost completely. Where London is mentioned at all it is incidental to some other main line of argument. 1 The point is well illustrated in the collection of essays edited by Abrams and Wrigley, Towns and societies, especially in Abrams's essay in the volume. 2 Adam Smith stressed the mutual stimulus that urban growth and agricultural improvement might afford each other. Smith, Wealth of nations, book iii, chapter 4, 'How the commerce of the towns contributed to the improvement of the country'. 157 Urban growth Urban growth and agricultural change economy. If productivity per head in agriculture is sufficiently low the surplus of food available after meeting the needs of the agricultural population may be enough to sustain only a tiny urban sector. At the other extreme, if agricultural productivity is high the economy may be able to support, say, a third or a half of the population in towns without prejudicing nutritional levels elsewhere. In a closed economy, therefore, a substantial rise in the proportion of the population living in towns, is strong presumptive evidence of a significant improvement in production per head in agriculture, and may provide an indication of the scale of the change. Sufficient is now known to justify an initial application of this line of thought to early modern England. quadrupled and the city's share of the national total rose from 2.25 to 5 per cent. The percentage of the population living in other towns, on the other hand, rose only modestly from 3 to 3.25 per cent. Even this was largely because ten towns crept over the 5000 mark (Plymouth, King's Lynn, Gloucester, Chester, Hull, Great Yarmouth, Ipswich, Cambridge, Worcester, and Oxford). Most large provincial centres were growing less quickly than the national population as a whole. This appears to have been true of Norwich, Bristol, Exeter, Canterbury, Coventry, Colchester and Salisbury. Indeed, of the initial list of towns above 5000 in population, excluding London, only Newcastle increased its share of the national total, and its surge in growth was no doubt partly due to London's extraordinary rise, since the coal trade down the east coast flourished as London grew and brought prosperity to Newcastle in its train.3 The effect of the artificial boost given to urban growth outside London by the inclusion of several new towns which had reached 5000 in population between 1520 and 1600 can be estimated, either by ignoring the new entrants on the 1600 list, or by basing the calculation on the full 1600 list. The group of nine provincial towns on the 1520 list displayed a collective growth of only about 15 per cent, sluggish growth during a period when the national total population rose by about 70 per cent. If, alternatively, the list of 19 large provincial towns in 1600 is made the basis of measurement, a rather higher percentage growth figure results. In 1520 they appear to have housed a total population of about 107 000.4 By 1600 the total had risen to about 137 000, a rise of 28 per cent. Both calculations underline the point that the doubling of population living in provincial towns shown in table 7.2 is misleading since it was preponderantly due to the recruitment of new towns into the category, and not to growth within the individual towns. If London's growth is ignored, the urban growth pattern elsewhere conforms to expectation quite well. In the course of the sixteenth century real incomes in England appear to have fallen quite substantially. If the Phelps Brown and Hopkins (PBH) index of real wages is used as a guide to the extent of the fall in living standards, it suggests a decline between 1520 and 1600 of over 40 per cent.5 Even if the PBH index overstates the change, a significant deterioration is none the less probable and ought, in conformity with the model sketched earlier, to have acted as a brake on urban growth and to have tended to reduce urban population in percentage 158 The pace of urban growth in England Table 7.1 sets out some estimates of the size of the populations of leading English towns in about 1520, 1600, 1670, 1700, 1750 and 1800: while table 7.2 provides estimates of the total population of England at each of these dates and of the population of London and of other urban centres with 5000 or more inhabitants. As the notes to the tables make clear, all the estimates given are subject to a substantial measure of uncertainty. Their sources are various and, apart from those taken from the 1851 census, most have been obtained by inflating and adjusting the raw data because the latter cover only a proportion of the total population. Some figures were originally suspiciously 'rounded' and may incorporate alterations made in the light of subjective assessments of their deficiences. Moreover, it is entirely arbitrary to draw a dividing line between the urban and the non-urban at a population of 5000. This was done on the supposition that only a tiny fraction of the inhabitants of towns larger than 5000 in population would have been principally engaged in agriculture, but a plausible case might be made for a significantly lower dividing line. To add still further to the crudity of the exercise, it is questionable whether the same dividing line should be used over a long period of time during which the population increased greatly. For example, a moderate-sized market town with a population of, say, 3000 at the start of the period serving a hinterland in which the population doubled, would itself grow in size and at some point exceed 5000 in population even though the functions it discharged did not alter. If this pattern were widely repeated, it would result in an upward drift in the overall 'urban' percentage, but this would have occurred only because the total population had risen, and it would not imply any structural change in the economy. Nevertheless, certain features of change over time are so prominent that they would remain clear cut, or might even be more pronounced, with better data. During the sixteenth century urban growth viewed relative to national population trends was largely confined to London where the population 159 3 The modesty of urban growth outside London may be overstated by concentrating on places with 5000 or more inhabitants. There is evidence to suggest that the smaller towns were growing more rapidly than the provincial centres. Using a total of 2000 rather than 5000 inhabitants to divide urban populations from the remainder would probably have resulted in an impression of greater buoyancy in urban population in the sixteenth century. PhythianAdams, 'Urban decay in late medieval England', table 1, pp. 171-2. 4 Using the sources listed in the sources note to table 7.1. 5 The construction of a slightly modified version of the PBH index is described in Wrigley and Schofield, Population history of England, appendix 9. The 25-year centred moving average of the annual figures given in table A9.2 fell by 44 per cent between 1520 and 1600. Urban populations (thousands) C.) Co C. o a) co CO zz c0 > ci C (f) (0(0 - (0 0 E cv C'J x >1 -C 0 0 (0 O).0 CJ (\J E Co D C (I) q> ('D c'J N- .c 'a) EOCO C\JE 0 -00 ,-D r- 0 n 0 (0 a) 0_i > LU (I, D C E LU L0 CO >.. >D Q)&5 - -(a a)0.0 0 0 a) >WLU . CODLU C000 01(0 CO -J 0(0 (!) Lo .0 C a)0 CQ) o o o 001 2 i _i .J Z 00- CO C/) CO )W '0 -0 E> D a) > E D -' C LucLooZc,)>-OOcLcC) 5 o .2- >.0 0 C 02 LU —i a)r~a)0oc030cOOxa)..QO0r IZCO 5OUJW J0C0)-OOCJ) ca)oCoLUD€ o Lo LU-C00 z co z w >- o m o C_(l) -J x.0 (0 CO (OOCOXO 00010 .2- 0 000 COO 0 .o LU LU LU LU c (C) C CO 5a) O> OLU.)a)LUa) a) z o w 0 .2-o (0 r-Co(0 (0(0 O) 00 C Ea) 11- 0 CD CO 0a)>.>'(0r C'.J (I) CO 0) C o -O 00 CD -g•fl: >- 0 O•C 0 a) -J z 0 a) - OD 5-5o LULU Nr- a) > o0o.a)xC-Q.00 -J Z >- cOZ W Lo Lo C\l CrY(0 0 00 c 0 FZ >< 0 0 a) (0 _iZCU>-COWOOZO Notes to table 7. 1 tD id cu) E c C: h 0 o a) 0) C -• 0 m cC °2 O'- (0 Q).0 () Q° '-c '—' C0)0 c (C a) a) < C0).-C N- €a E 0C x a) 0 0 a) E-_. 0 0 CUD LU '— ONC'— CL ca 0C C — - E4 • C) 0. --0 h-U CL x CL a)COCQ .cO.) 0 0 -j o C'J c > - 0 CL D .0(0 WCOCN-- C') S -5 c0 -C Ct) gC.. ig — 00 V) 0 -2- cri cEa) w.C°-'—O (D CD C COC/)00 0 cD'- 0 C '- '-5)2 0 , CSJ(O,j C) jo . Co.-Z 0 0 ci - 0(0 Caa E 0) ja)$ co 56 8 -'= 0000 cD0 ED a'- '-- 6E CC(0 co E m I- . - Q(1)C>, cc 0 — CQ) E - °> 0)Q.a Q)C 2ci C'JCJ 0 E 0 2">E (Ou•a) -.--' (000 Q) wj.. 0 •2 . COO QL E - -Q CU - o U, C .•.$ 0 00 0. E '•;CO >' 1 c cáO) CL 0 c 0(0 Eo ' d L ?C 2 Lo S C_.(o'- O 0(0 ' Ø0J0)° 0) E -.0 0Ow0. > soc-or',_O0Q)C0. a) OC t5 co • 000 0 E '-0°00 _>Q)•-_o g) g) gD - 55D25 Urban growth and agricultural change 0) . 00 00 C\1 00 -(') 0U)U) (D (D (0 (3) CU 2 a)o -U'-E a)° °w50 a) CE 2 CU W 0 0 L 0U) U)U)O r- rN-tO U)C\J U) - DZ -o (0 OD Ccia) 0O) - 'O' LO 0 C• 2 a 0)Qa) •U).N- C,) .?0~ 'U CU .2 g). 0D2-iiii I0 U) U)0 N- U) N0 U) CU Co U) U)L0 U)N- (0 oi 2cci c 0 wa C0 o2 C • - °-o a',- a)-2 - 2ii 6 75 2ag U)QU) O)C') U) 0 0 0 CJ(0 Nd -•-— °,_E.c'J ° -oa) i 0 .dE Ca O) 0) cci•18 w 5L National, London and other u rban population estimates (thousands) 2C0c_j o—a aai'- Ca OD '-.j2 . U') L0 00 -0 C')(') - (Y) j-CU U)U) QC'JC'J 0 U) 0 U) CU U) L0 CY) 00 0U) r-_ C\1 U) CU 0 CU c\.ic'-iLri 5 -ca 0 CL a Co (ci -o -o . q'. 0 ci) E or-'C _ _Q0) - 0. Cx —0 0) Caa) a) Q) D-a) - 2a 2ca Caa) 0 •>0 CG) 0 QCa QQ)()' E0-°.0 0 _ a) LO N 0) C a) '- O C 2a 0 .a) 2— o— Eã Ccia 2 • ' ._: >S a) a)G) (D D5 00 WC1 0) I-C - 4c 2g 0) C o .20) 0 C o C. Ca 0Q)Ca DC0 CoE D -9D0) CC0 c C0 0 WiO H 0aC5 N(0U2 ci) 0 U) -c '29 — U -0 - .'5? (D 0 cO> >c5W.oEco Ccir. -.?OOci) •QCC)Q o j C LO a) a O .0)0) 0) ( o - i2 0 (1) 0) terms.6 London remains as an exception so important as to outweigh in aggregate faltering urban growth elsewhere, but its overall dominance should not be allowed to obscure the significance of trends in the provinces. In the seventeenth century circumstances changed greatly. The national population grew by less than a quarter over the century as a whole, and was falling gently during its third quarter. The PBH index suggests that real wages bottomed out early in the century and had risen substantially by its end.7 Urban growth went on apace whether judged in absolute or percentage terms. London continued to dominate the picture. By 1700 the capital housed perhaps 11 per cent of the total national population, more than double the percentage of a century earlier. It had become the largest city in western Europe and continued to dwarf all local rivals. But other towns also began to grow vigorously. They grew rather slowly in the first half of the century, but after 1670 their relative growth was at least as rapid as that of London. And the smaller urban centres were now increasing far faster than the country as a whole. Their population more than doubled during the century, a rate of growth more than four times that of the national aggregate. Measurement of urban growth is less bedevilled in the seventeenth century than its predecessor by the problem of 'drift' across the arbitary 5000 dividing line between urban and rural. In the sixteenth century, as we have seen, the increase of population in towns of 5000 or more people was about 95 per cent judged crudely, but the increase in the towns represented on the 1600 list was only 28 per cent, while the increase in the towns on the 1520 list was only some 15 per cent. The comparable figures for the seventeenth century were more closely bunched at about 105, 60 and 46 per cent respectively.8 Several major provincial centres, notably Norwich, Bristol, Newcastle and Exeter, increased by between 50 and 100 per cent. A striking portent for the future was the appearance on the list for the first time of towns never previously of much note but later to herald a new age. In 1670 Birmingham, Manchester and Leeds appear for the first time, and in 1700 Liverpool. If the seventeenth century saw a notable acceleration of growth within an urban system still consisting largely of long-familiar names, the eighteenth brought a radical re-ordering of the urban hierarchy and further rapid urban growth. London, moreover, though still vastly larger than any other city, ~C CD Co C W 2 a) 0. Ca 0) Ca (,) C CU 0 20 CaO UJ c CaC O co Q.CC2 C.Q0 CaDa) .00C-cC3 0-0 D 163 -J OH C 0 a) . C') - LU 2 .iIc•a0) 3 a) • w (ci.. - CaCo I- 0= 0 C -CO C 0, < IL Cci . I- Ct) 00 a) 6 For a recent survey of this and cognate issues see Palliser, 'Tawney's century', especially pp. 349-51. Palliser suggests reasons to suspect that the PBH index overstates the extent of the fall in real incomes. 7 The 25-year centred moving average suggests a rise of 27 per cent between 1600 and 1700. Both the extent of the rise and the timing of the end of the long fall in Tudor and early Stuart times are debatable. Bowden's index of the purchasing power of agricultural wages in southern England suggests that the beginning of a recovery may have been as late as the 1640s. From a stable plateau in the 1520s and 1530s (the index figure is 80 in both decades where 1460-9 = 100), his index falls by 35-40 per cent by the 1590s and shows no subsequent decisive trend until the series ends in 1640-9. Bowden, 'Statistical appendix', table xvi, p. 865. 8 In order to calculate the middle figure (60 per cent) some town populations had to be estimated (i.e. for some of the towns appearing on the 1700 list but not the 1600 list in table 7.1). Urban growth Urban growth and agricultural change no longer stood out for its rate of growth. In 1801 it comprised much the same proportion of the national population as a hundred years earlier. Meanwhile, the share of other towns larger than 5000 in population increased sensationally, rising from 6 to 17 per cent of the national total, and for the first time surpassing London's share. Growth was widely but very unevenly spread. London's old rivals fared less well than London in the main. Bristol grew rapidly, riding on the back of buoyant Atlantic trade, but several cities which had once figured prominently in the English urban hierarchy grew less quickly than the population overall and ended the century with smaller fractions of the national total than at the beginning, including Norwich, Exeter and York. For many centuries such towns had exchanged places in the premier urban league but the league's membership had not greatly altered. By 1801, however, only Bristol, Newcastle and Norwich of the old major regional centres remained amongst the country's ten largest towns. Manchester, Liverpool and Birmingham stood second, third and fourth after London. They ranged between 70000 and 90000 in population, having grown fifty-fold or more since the early sixteenth century. Lower down the list a host of new names appeared. Several were the seats of new industry - Leeds, Sheffield, Stoke, Wolverhampton, Bolton and Stockport - but others reflected changing social customs and new forms of expenditure. Bath, for example, was the twelfth largest town in England with 33 000 inhabitants, a gracious monument to the new ways in which the wealthy and well born found it convenient to make or maintain contacts with each other or to pass their hours of leisure. Ports and dockyard towns also enjoyed vigorous growth. Plymouth and Portsmouth were among the country's largest towns, and Hull, Sunderland, Chatham and Great Yarmouth all exceeded 15000 in population at the time of the 1801 census (though in this ports category not all the towns were new names). The simplest model connecting real income and urban growth will no longer 'save the phenomena' for the eighteenth century. The sustained momentum of urban growth, accelerating towards the century's end, would suggest a parallel rise in real incomes, but in some parts of England the long-sustained rise in real wages had ceased before the middle of the eighteenth century and it had probably halted nationally by 1780, to be succeeded by a declining trend lasting perhaps 30 years.9 By the eighteenth century, of course, the assumption of a closed economy is even more unrealistic than for the sixteenth. External demand represented a substantial fraction of total demand in many industries, though it is easy to exaggerate the importance of overseas markets)° Any increase in the relative importance of overseas trade, however, would stimulate urban growth, conspicuously in the case of ports like Bristol and Liverpool, but in a lesser degree also elsewhere. Equally, transport improvements within England increased the scale of internal trade. The average distance travelled by goods between producer and consumer probably also increased. Both trends must have stimulated employment in the urban foci through which goods passed.11 The fortunes of different types of cities suggest that the deceleration in urban growth which might have been expected on the simplest possible view of the link between real incomes and urban growth affected the older centres in the expected fashion, but that the new features of the English economy imparted impetus to those towns most caught up in the new developments. London, affected by all the various and conflicting influences on urban growth, occupied an intermediate position. The extent of the contrast between the fortunes of different types of towns in the course of the early modern period is illustrated in table 7.3. The choice of towns in each group is inevitably arbitrary, both in the sense that other sets might have been made up to represent the type in question, and in the sense that large towns are unlikely to conform to 'pure' types since size and complexity of function are closely linked. For example, the balance of market functions, craft industry, administrative services and professional employment varied considerably among the set of ten historic regional centres, and two of them, Chester and Exeter, also had important port functions. Nevertheless, the contrast between the groups is sufficiently marked to make it unlikely that other choices made with the same distinctions in mind would produce a very different result. The historic regional centres did not keep pace with national population growth over the early modern period as a whole. In the two middle periods, when real wages were rising, they experienced a faster population growth than the national average; on the other hand, in the sixteenth and later eighteenth centuries, when real wages were probably declining, their rate of growth fell well below the national average. Thus, they exhibited what might be termed the 'classic' pattern of relative growth in terms of the model of the relationship between income and urban growth described earlier. At the other extreme, the four towns which in 1801 were the largest manufacturing towns in England were always growing faster than any other group in the table, apart from London down to 1700. Their rate of growth accelerated steadily throughout the three centuries, becoming so hectic in the last half-century that their population almost quadrupled in 50 years. The established ports also grew with increasing speed except in the last period, outstripping the national growth rate except in the first and last periods. The slight fall in growth rate in this group in the later eighteenth century was no doubt due in part to the extraordinarily rapid growth of Liverpool. If Liverpool were included in the group the percentage growth of population in the group would rise to 72 in 1700-50 and 82 in 1750-1801. 164 9 The behaviour of wages and prices, and a fortiori of real wages, both regionally and nationally in the period 1700-1850 has been the subject of much controversy. There have been several valuable surveys of the issues, and also some new empirical work. Flinn, 'Trends in real wages'; Von Tunzelmann, 'Trends in real wages', Lindert and Williamson, 'English workers' living standards'. 10 See, for example, Thomas and McCloskey, 'Overseas trade and empire'. 165 11 An excellent summary of knowledge about these topics may be found in Chartres, Internal trade in England, especially chapters 2 and 3. Urban growth and agricultural change oc'j Percentage growth LO j— 00 co 0) -L() IL) (Y) OD C\1 00 N- C') LL) U) N-(O coo) — Cl) (') C') 0 U) 0) c'J (0 c'i c-'J 00 (0(0 (00) (0 London contrasts sharply with each of the other three groups, growing far more quickly than other groups until the end of the seventeenth century, even though it became so large that in 1700 London housed more than twoand-a-half times as many people as the other three groups combined. Thereafter, however, London was outpaced by both the ports and the 'new' manufacturing towns, and did not even quite match the national average growth rate. The foregoing is both compressed and simplistic. Uncertainties of definition, estimation and periodization have been dealt with summarily or ignored. Nor has the nature as opposed to the quantity of urban growth been explored. It is possible, for example, that much of the growth of sixteenth-century London was a 'push' phenomenon linked to the scale and depth of rural poverty, and due to what Clark termed 'subsistence' migration; whereas later, in contrast, movement to the capital may have had a greater 'pull' element as living standards rose and 'betterment' migration came to predominate.12 It remains reasonable to argue, however, that there were important links between some types of urban growth and real income trends in early modern England, though equally clear that urban growth is not to be explained solely in this way. Neither the headlong growth of London in the sixteenth and early seventeenth centuries, not the acceleration of urban growth in the 'new' manufacturing towns in the later eighteenth century is explicable in terms of the behaviour of domestic real income per head, yet both were developments of massive importance. (\J-- 00 co 0) Cci ci) 0 0 (ci .0 Population total Urban growthin early modern England (thousands) d 0 L N- N. N. (0 U) (0(0 c-,J c\J 0 N. 167 cE Ct( ci).0 .0 3 CC ci) >- (.5 0 U) (0 N0 U) U) N. 0(0 05C -o E Co 00 '-0 (.'j Urban growth and agriculture > .J 0 Q, C') C') -C 3 E iii> r- LO Xci) UJ5 ci ci) (I) oiO c'.jco OIL) (DC') ci (N (0 o 0. - N ci) ci)•Vi uLO Cl) occi 0)0 0) 'DO -c o.Z .a >'..: 3 -0 (I) t1 a) ) C Q) ca >-5 CO (n - I-C.. O)(d) znm, Even if the causes of urban growth are elusive, the fact of growth remains, and some of its implications can be examined. In any pre-industrial community agriculture is the dominant form of economic activity and the levels of productivity per head set in agriculture necessarily govern the growth opportunities of other industries. This was a point so well known to those living in pre-industrial economies as scarcely to warrant remark, and when political economy reached its first great statement in the Wealth of nations Adam Smith made the examination of this issue one of the chief concerns of the work. 13 But agricultural productivity has proved very difficult to 12 Clark, 'The migrant in Kentish towns'. 13 Smith's chapter, 'Of the different employment of capitals' includes a strong plea for agricultural investment as the ultimate basis of national productive capacity. 'The capital employed in agriculture, therefore, not only puts into motion a greater quantity of productive labour than any equal capital employed in manufactures, but in proportion, too, to the quantity of productive labour which it employs it adds a much greater value to the annual produce of the land and labour of the country, to the real wealth and revenue of its inhabitants. Of all the ways in which a capital can be employed, it is by far the most advantageous to the society.' Smith, Wealth of nations, i, p. 385. Urban growth Urban growth and agricultural change measure directly. One way of measuring it indirectly is to consider the extent of the rise in agricultural productivity suggested by the course of urban growth in England, taking into account also changes in the occupational structure of the rural component of the population. To simplify the calculation, I assume that consumption of food per head did not vary between 1520 and 1800 and that England was neither a net importer or a net exporter of food. The first assumption is doubtful and the second is demonstrably false, especially during the eighteenth century. In its early decades England was a substantial net exporter of grain, and towards the end of the century substantial quantities of meat and grain reached the English market from Ireland. 14 It is, however, convenient to begin with such assumptions. In 1520 the urban percentage was 5.25: in 1801, 27.5 (table 7.2). This in itself suggests a useful gain in agricultural productivity. In 1520, 100 rural families fed 106 families in all; in 1801, 138 [100 x 100/(100 - 5.25) = 106; 100 x 100/(100 - 27.5) = 138]. The level of productivity in 1801 is 30 per cent higher than in 1520, far from a negligible increase, if scarcely sensational. But any such exercise must understate the extent of the increase in agricultural productivity if there is also a decline in the proportion of the rural labour force engaged in agriculture. There can be no reasonable doubt that such a decline occurred. In certain rural areas in the eighteenth century, indeed, the growth in non-agricultural employment was so great as to dwarf the remaining agricultural population. Framework knitting became the dominant source of employment in many Leicestershire villages. In parts of Warwickshire and Staffordshire there was very rapid growth in the manufacture of small metal wares: nails, chains, buckets etc. In much of Lancashire and the West Riding of Yorkshire the textile industry, whether cotton or wool, provided income for many more men and women than did agriculture. The steady growth of coal production in Northumberland and Durham produced the same result in substantial tracts of these counties. Even in more strongly agricultural counties in the south, lace-making, straw-plaiting and the like provided much employment for women. Moreover, in areas which attracted little industry there was often a substantial growth in employment in service industries. In the rare cases where parish registers provide data on occupation over long periods of time it is a commonplace to note a growth in specialist employments not previously encountered, especially during the eighteenth century. A small town like Colyton in Devon, for instance, even provided a living for a peruke-maker in the 1760s. Almost everywhere the proportion of men described as labourers, husbandmen, yeomen or farmers tended to decline as a proportion of all the occupations mentioned. It is true that many craftsmen also owned scraps of land and its produce was of crucial significance in their domestic economy. Even those who worked no land might nevertheless be drawn into the labour of harvest. However, the reverse was also often true. Those to whom an agricultural occupation was attributed might turn their hands to craftwork during the seasonal slacks on the farm. Ideally, it would be preferable to measure hours worked in different forms of employment rather than to treat each member of the workforce as uniquely engaged in a single occupation, but it is enough for the present purpose to show that there was a major fall in the proportion of the rural labour force in agricultural occupations. By 1801 a tolerably accurate picture of rural employment structure can be drawn. Deane and Cole estimated that 35.9 per cent of the labour force in 1801 were engaged in agriculture, forestry and fishing.15 If we assume, for simplicity's sake, that none of this occupational grouping lived in towns, then it follows that only some 50 per cent of the rural population were engaged in agriculture given that the rural population comprised 72.5 per cent of the whole (35.9/72.5 x 100 = 49.5). The comparable figure at earlier periods is unfortunately difficult to establish, though useful clues may be found in the work of King and Massie. In order to make revised estimates of the changes in agricultural productivity, I have assumed that 80 per cent of the rural labour force were engaged in agriculture in 1520; that this figure declined very slowly to 70 per cent by 1670 with the bulk of the fall occurring after 1600; more quickly to 66 per cent in 1700; and then linearly in the eighteenth century to 50 per cent in 1801. 16 These assumptions allow the population to be subdivided into three groups rather than two; the urban population, the rural population engaged in 168 14 Brinley Thomas has recently estimated the relationship between the value of imports of grain, meat and butter and the income of British agriculture. In 1814-16 total imports of the three commodities represented 6.4 per cent of British agricultural income, and of these imports 70 per cent came from Ireland. See Thomas, 'Escaping from constraints', table 2, p. 743. Jones recently estimated that 90 per cent of the population of Great Britain were fed from domestic agricultural production in 1800. Jones, 'Agriculture, 1700-1780', p. 68. In contrast, in the first half of the eighteenth century English net grain exports were a substantial fraction of total production, reaching perhaps 6 per cent of gross domestic grain output about the mid-century. Deane and Cole, British economic growth, table 17, p. 65; also Ormrod, 'Dutch commercial and industrial decline'. 169 15 Deane and Cole, British economic growth, table 30, p. 142. Alternative estimates of the scale of employment in early nineteenth-century agriculture may be found in Wrigley, 'Men on the land', especially table 11.10, P. 327, and accompanying text. 16 The first serious attempt to quantify the occupational and social structure of England before 1801 was that carried out by Gregory King in 1688. When Deane and Cole considered his estimates, they suggested that 'between 70 and 80 per cent of the occupied population was primarily engaged in agriculture'. Deane and Cole, British economic growth, p. 3. This is implausibly high. Having regard to the total national population in 1688 and 1801 (4.90 and 8.66 million) and the proportion in agriculture (say 75 per cent and 35.9 per cent), such a high proportion in 1688 would imply a fall in the absolute scale of agricultural employment of more than 15 per cent over the intervening century. Lindert has recently re-examined the changes in English occupational structure in the eighteenth century but the results, at least for agriculture, do not appear convincing. The percentage of the male labour force employed in agriculture is estimated at 22.3 in 1700, 26.1 in 1740 and 13.7 in 1811. Lindert, English occupations, table 6, pp. 46-7. See also note 19 below. Urban growth and agricultural change o ccU) (Y) cc (D 0) o o (N (00 (N (N (') U) (NO) (0 - N-C'4(N (') -'a- (occ 19 U) 00) C') '- cc cc cc 000000 NN- 0 000000 (NU)U)(00 N- (N 0 - U) ci) cc CU N. - q, ccccoN- co - -o cc'i ca 0)U)c0o Pcc(N(occc')(0 0(NC')C') t0)-(0--0 '-C'J(00 0C')'- - ICT co 16 0(00(0 cc 0 a) cx: aN-r-LoLr) 0000 d d D (0 U) .20OU)00(N 0(000U) (0 (0 0)')N-N-(0U)C') a) N- r- 0 '- U) CO (N N- C') (N U) (N - cc- (oo) - o co (0U)(N U) O) .-.-Lr)0)0)cccco 0U)0O)O)- 0 0 X OD E i5 a. cc cc U) (N (c0:C\CV) 0000'-(N (') It 0 - cc co N- (0 L(U)cO Q0U)OQU) .9U) U)ccc')N--N- -U)0)(0-0 (N (N 0 '- (N C') U) 0 75 cc 0 cc 0 76 a) c 0 0 - - C 0 co 000000 a.(N0N-0U) a)U) (0 (0 N- N-- a. U) ) Ca c 00000W(N 0 N- 0 U) 0 U) (0 (D N- N- -0 (NN-N- Nr U) c (00) (N 0) CL (N0N-0U)0a.c\J0N-0U)0 -0N-0U)0 0 U) (o (c) N- NU) (0 (0 N- N(0(0 N- N- cc CL cc cc Source: For population totals in top panel see table 7.3 and discussion in text. Urban, rural agricultural and rural non-agricultural populations (millions) I- 171 agriculture, and the rural population dependent on employment other than in agriculture. At the same time changes in output per head in agriculture can be calculated. The results are set out in table 7.4. Table 7.4 suggests that the rural agricultural population scarcely changed in size for a century and a half between 1600 and 1750 and that even in 1801 was only a tenth larger. than 200 years earlier.17 It is not therefore surprising that table 7.4 should show a striking rise in agricultural productivity, and indeed to the degree that the assumptions that have been made in constructing the table are justified the conclusions are inescapable. 18 The crucial assumptions are that England was not a significant net importer or exporter of food; that the growth of the percentage of the population living in towns followed the pattern set out in table 7.2; and that the proportion of the rural population engaged in non-agricultural production rose from 20 to 50 per cent between 1520 and 1801. For the second and third assumptions, the central issue is the extent of change over the period as a whole; the question of its timing, though fascinating, is of lesser importance. Precision about any of the three basic assumptions is beyond reach. On the first it is worth noting that the fact that England was a substantial food exporter in the early eighteenth century means that agricultural productivity is probably understated at that period. Equally, by the end of the century England had become a net importer, especially from Ireland, which implies the opposite. The second set of assumptions concerning urban growth is probably sufficiently accurate to avoid significant error. The third assumption relates to the agricultural proportion of the rural population. The figure for 1801 is fairly firmly grounded in the evidence of the early censuses but earlier estimates become increasingly fallible. Any figure for the early sixteenth century must be largely guesswork. A lower figure would, of course, reduce the apparent gain in agricultural productivity, but only a radically lower figure would greatly change the general picture. 19 17 Jones estimated that the agricultural population of England and Wales increased by 8.5 per cent in the eighteenth century. Jones, 'Agriculture, 1700-1780', P. 71. 18 Crafts's calculations produce results broadly similar to those implied by table 7.4 in relation to agricultural productivity. He estimated that agricultural output rose at the following percentage rates per annum over the periods 1710-40, 1740-80 and 1780-1800: 0.9, 0.5 and 0.6. This implies a total increase in output of almost exactly 80 per cent over the 90-year period as a whole. Assuming that the agricultural labour force grew by 13 per cent over the century, as table 7.4 suggests, the increase in output per man, at 59 per cent, is closely similar to that which may be calculated from column 7 of table 7.4 (52 per cent). Crafts, 'The eighteenth century', tables 1.1 and 1.2, pp. 2, 3. 19 It is possible to use the work of Gregory King and of Joseph Massie to test how far the estimates made by contemporaries agree with those in table 7.4. In 1760, out of a total of 1 472 000 families, Massie supposed that 210 000 were freeholders, 155 000 farmers and a further 200 000 husbandmen, making a total of 565 000 unambiguously engaged in agriculture. In addition, he estimated that there were 200 000 families of labourers outside London. Not all of these Vc'ould be agricultural labourers. Massie reckoned that there were 20 000 labourers in London, only slightly short of the number to be expected on the assumption that labourers were as numerous per 1000 population inside cities as in the countryside. We may assume, Urban growth Urban growth and agricultural change Other assumptions could, of course, be used in constructing table 7.4. For example, if one were to assuitie that 15 per cent of English food requirements were met by imports in 1801 and that agriculture employed 55 per cent of the rural labour force rather than 50 per cent, while other assumptions were unchanged, the estimated overall rise in agricultural productivity would be reduced from 109 to 61 per cent. Going still further, if the 1520 figure in column 4 were reduced from 0.80 to 0.70 and the modified assumptions for 1801 were retained, the figure would drop even more to 41 per cent, but this figure is improbably low.20 On present evidence, therefore, while the particular figure given in table 7.4 is, of course, arbitrary, there is a strong likelihood that the 'true' figure lies between 60 and 100 per cent. The phasing of changes in the proportion of the rural labour force engaged in agriculture no less than their scale is also largely a matter of judgement rather than a demonstrable pattern. The same pressures which kept urban growth outside London to such modest proportions in the sixteenth century are likely to have restricted employment opportunities outside agriculture. It therefore seems appropriate to make only a small reduction in the proportion of the rural population in agriculture between 1520 and 1600. Thereafter, the pace of change probably accelerated. One further implicit assumption deserves discussion. Individual intake of food measured in calories varies within fairly narrow limits. There is little evidence of widespread malnutrition so extreme as to cause death in early modern England. With rare and usually local exceptions, even severe harvest failure did not provoke heavy mortalities.2 ' It is improbable (though also undemonstrable) that mean personal daily calorie intake varied in a manner which would significantly undermine the line of argument deployed above. Nevertheless, periods of rising real income must have been periods in which per caput food consumption tended to rise with the opposite happening in times of declining living standards.22 In addition, there were changes in the composition of individual diet as incomes rose or fell. Meat was a luxury to the pauper but a commonplace to more prosperous members of society, and there were secular shifts in the relative prices of grain and meat which reflected the long-term trends of the average real income.23 Inasmuch as food such as meat and dairy produce needed larger inputs of labour, as well as land and capital, to produce a given number of calories of food compared with grain, it might seem that an allowance should be made for the impact of secular real income trends in attempting an individual measure of agricultural productivity. This in turn would imply that agricultural productivity per head was rising faster during the seventeenth and early eighteenth century than suggested by table 7.4, but less quickly during the later eighteenth century. It also suggests that it may have been falling in the sixteenth century, a finding in keeping with common sense since rural agricultural population is estimated to have risen by almost 60 per cent between 1520 and 1600, a scale of increase very likely to involve a falling marginal productivity of labour and much concealed or overt underemployment. Making an explicit allowance for real income changes, however, presents problems. At present, real income data are based on slender foundations and involve wide margins of uncertainty. Little is known about any changes that may have occurred over time in the income elasticity of demand for 172 therefore, that another 15 000-20 000 were to be found in other urban centres, and it is probable that a further group, though living in the countryside, was employed outside agriculture, especially in the building trades (which were not separately itemized by Massie). Assuming that these, too, were about 20000 in number, then the overall total of those families engaged in agriculture would be c.730 000, or fractionally less than one-half of the national total. The comparable figure in table 7.4, that for 1750, is 46 per cent, a broadly similar figure. A comparison with Gregory King (1688) is more difficult to make because of the form in which he drew up his famous table. King's totals for families of freeholders and farmers are not greatly different from Massie's (180 000 and 150 000 respectively) but he gave no separate figure for husbandmen. Instead, all other families which might be engaged in agriculture appear in two categories, labouring people and outservants (364 000) and cottagers and paupers (400 000). It is clear that most of those engaged in industrial crafts - weavers, glovers, knitters, tanners, carpenters, coopers, sawyers, thatchers, smiths, wrights, cordwainers and so on - must have been included in one of these two categories (King lists only 60 000 families of artisans or handicraft workers in the entire country), together with all labouring families living in London or other towns. If we assume that half of the cottars were what Massie would have termed husbandmen and half engaged in industrial crafts, a split similar to that found in Massie's separate tabulation of the two categories; that 15 per cent of all labourers were in towns (where at this date about 15 per cent of the population lived); and that of the remaining 309 000 labourers, 25 000 were employed outside agriculture, the total of families employed in agriculture is 814 000 or 60 per cent of the national total of 1 361 000 families. The comparable figure in table 7.4 is about 58 per cent (1670, 60.5 per cent; 1700, 55.0 per cent). There is a convenient reproduction of both King's and Massie's tables and a discussion of the difficulties in comparing them in Mathias, The transformation of England, pp. 171-89. 20 Sixteenth-century estimates of occupational structure are inevitably very insecurely based, and it is possible, even likely, that the dominance of agriculture is too easily overstated. The muster roll for Gloucestershire excluding Bristol taken in 1608, for example, suggests that only 46.2 per cent of the adult male population aged between 20 and 60 years were engaged in agriculture. The comparable figure when the 1811 census was taken was virtually the same at 45.8 per cent. The economic history of Gloucestershire is far from typical of the country as a whole, but Gloucestershire affords an example which should caution us against assuming too readily that rural non-agricultural employment was always very limited in Tudor or early Stuart England. Tawney and Tawney, 'An occupational census'; 1811 Census, Enumeration, Parliamentary Papers 1812, xi, p. 121. 173 21 Wrigley and Schofield, Population history of England, especially pp. 320-40, 370-82 and appendix 10. The principal exception to the rule that harvest failure did not provoke big mortalities was the north-west of England in the sixteenth and early seventeenth centuries. Appleby, Famine in Tudor and Stuart England. 22 Estimating income elasticity of demand for food in eighteenth-century England presents great difficulties for lack of relevant data. In a review of the limited evidence, Crafts concluded that the most probable figure for the late eighteenth and early nineteenth centuries was 0.7. Crafts, 'Income elasticities of demand', especially pp. 154-9. 23 Kussmaul provides a convenient graph of the relative prices of the two commodities. It bears a strong resemblance to an inverted graph of the Phelps Brown and Hopkins real wage series, though the match is by no means perfect, especially in the early seventeenth and late eighteenth centuries. Kussmaul, Servants in husbandry, figure 6.3, p. 104. Real wage data based on Phelps Brown and Hopkins are set out in Wrigley and Schofield, Population history of England, appendix 9, and shown graphically in figure 10.5, p. 414. 174 Urban growth food. It may be noted, however, that the only existing real wage series covering the whole period, that of Phelps Brown and Hopkins, stood higher in 1800 than it had two centuries earlier.24 The apparent gains in agricultural productivity over the seventeenth and eighteenth centuries as a whole are therefore unlikely to be overstated because of a failure to take real income explicitly into account. Labour released from agriculture is available to increase other forms of production. The gross changes were striking. The rural agricultural population was 76 per cent of the total population in 1520 but only 36 per cent of the total in 1801 (table 7.4), and non-agricultural employment therefore grew from 24 to 64 per cent of the whole. This may overstate the extent of the change in that some of the growth of non-agricultural employment represented jobs created by increased specialization of function. A carter, for example, making his living by moving to market goods previously taken there by local farmers may be placed outside agriculture in an occupational breakdown but undertakes a task previously performed by the farmer. Yet the change was great. 'Adam Smith argued that a surplus of agricultural production over the food needs of the farming population might either be consumed unproductively by, say, retinues of servants, or productively if the surplus maintained an army of 'manufacturers' whose output added to the wealth of the community as a whole.25 In early modern England the growth of employment in industry and commerce is a testimony to the predominantly 'productive' use to which the growing relative surpluses in the agricultural sector were put. Adam Smith considered that the scale of such growth was largely conditioned by the extent of the rise in agricultural productivity. 26 He did not envisage the much more radical type of change that has come to be called an industrial revolution, nor is there any compelling reason to suppose that even increases in agricultural productivity as striking as those achieved in England between 1600 and 1800 will necessarily engender an industrial revolution. Yet the scale of change in early modern England bears stressing. It stands out more clearly if comparison is made with other countries. Urban growth on the continent De Vries has undertaken a very informative analysis of urban growth patterns between 1500 and 1800 for Europe as a whole and for some major regional 24 The index stood at 409 on average for the 11 harvest years 1595-6 to 1605-6, compared with 507 in the period 1795-6 to 1805-6. Wrigley and Schofield, Population history of England, table 9.2, pp. 642-4. 25 Smith's chapter 'Of the accumulation of capital, or of productive and unproductive labour' deals with this topic in the Wealth of nations, book ii, chapter 3. 26 The passage leading up to the assertion that 'the manufacturers of Leeds, Halifax, Sheffield, Birmingham and Wolverhampton . . . are the offspring of agriculture' exemplifies his argument very well. Smith, Wealth of nations, i, p. 431. Urban growth and agricultural change 175 I 1500-1600/1650 II 1600/1650-1750 III 1750-1800/1850 Rank Figure 7.1 Urbanization in Europe. Source: De Vries, 'Patterns of urbanization', figure 3.5, p. 97. subdivisions.27 In it he made use of an empirical relationship between the sizes of the towns and their position in a rank ordering whereby the difference in their populations is proportional to the difference in their rank orders. Thus p1 = p1 /1 where Pi is the population of the largest town and i refers to the rank order of a town after all towns have been arranged in descending order of size. The population of the fifth largest town may therefore be expected to be a fifth of that of the largest, and so on. If logarithmic scales are used to plot the population and rank coordinates of each town, the resulting distribution in the archetypal case will fall on a straight line with a slope of 45 degrees. In practice the angle of the line may vary somewhat from the 45° slope suggested by the strictly proportional relationship, and the first few points plotted are sometimes aberrant, as when the largest settlement is very much bigger than the second largest. But regularities in town size distributions are often impressive, and any anomalies within a data set, or changes in the angle of the plotted slope over time, may be illuminating. By compiling rank-size date for European towns at intervals over a threecentury period starting in 1500, de Vries was able to establish changes in the slope and shape of the urban hierarchy sufficiently pronounced and consistent to distinguish three major periods, as shown in figure 7.1. At the beginning of the sixteenth century the slope was gentle and there was a flat 'top' to the distribution. De Vries attributed the latter to the still strongly regional character of the European economy which was insufficiently articulated to produce leading urban centres of the size implied by the slope of the lower part of the distribution. During the sixteenth century the rank-size 27 De Vries, 'Patterns of urbanization'. Urban growth Urban growth and agricultural change Table 7.5 Urbanization in England and the continent: percentage of total population in towns with 10 000 or more people Table 7.6 Urbanization in England and the continent (revised): percentage of total population in towns with 10000 or more people 176 1500 England North and west Europec Europed aC. 3.2a 6.0 6.1 1600 1650 1700 1750 1800 6.1 8.1 8.0 108b 13.4 13.0 9.5 17.5 13.8 9.9 24.0 14.7 10.6 10.7 9.3 1520. bC. 1670. 'Scandinavia, The Netherlands, Belgium, England, Scotland, Wales and Ireland 'The countries included in Europe are Germany, France, Switzerland, Italy, Spain and Portugal together with those listed as north and west Europe. Sources: For England see source notes to tables 7.1 and 7.2. Other data are from de Vries, 'Patterns of urbanization', table 3.6, p. 88. plot gradually straightened. The continent-wide economy was becoming more integrated, and cities such as London, Paris and Amsterdam grew very rapidly as they assumed urban functions over wide hinterlands. A second period then supervened, running from about the early seventeenth to the mid-eighteenth century. In general, in this period, the larger the town, the more rapidly it grew, so that a disproportionate part of the overall rise in urban population took place in the larger towns. The rank-size line pivoted slowly round a point close to its lowest reading and thus grew steadily steeper. After about 1750, however, urban growth changed in character again and a third period began. The rank-size line moved outwards from the origin of the graph implying a rise in the number of towns in all size categories, but there was a more rapid growth in the number of smaller towns than of larger ones. The slope became less steep once more. The sequence of changes was shared by the major subdivisions of Europe used by de Vries, though the timing of the changes varied somewhat. The pattern of change in England, however, was quite unlike that on the continent. As table 7.5 shows, England began the period with an unusually slight proportion of her population living in large towns, but jassed the European average in the mid-seventeenth century and the north-west European average by 1700. By the beginning of the nineteenth century England was relatively heavily urbanized. Events in England, however, were so different from those elsewhere as to distort patterns of change when England is included in some larger entity. It is instructive to remove England both from the urban and the overall population totals used to generate urban percentages for continental areas. The result is shown in table 7.6. Urbanization in north and west Europe recedes rather than advancing substantially in the eighteenth century, though the scale of advance in the seventeenth century is not greatly altered. In Europe as a whole the exclusion of England slows the increase in urbanization in the seventeenth century and brings it almost to a halt in the eighteenth. 28 28 There might be reason to suppose this finding spurious if de Vries's estimates of the population of English towns and those shown in table 7.1 were widely divergent, but a detailed comparison of his data with those of table 7.1 suggests that this is a groundless fear. England North and west Europe minus England Europe minus England 177 1600 1700 1750 1800 6.1 13.4 17.5 24.0 9.2 8.1 12.8 9.2 12.1 9.4 10.0 9.5 De Vries provides estimates of population totals for north and west Europe and for Europe as a whole for 1600, 1700, 1750 and 1800 in the work listed below (though not for the other dates in table 7.5). This information and the urban percentages given in table 7.5 make it possible to calculate the size of the urban populations. The English totals can then be removed both from the urban and total populations and the percentages recalculated. Sources: For England see source notes to tables 7.1 and 7.2. For other data see source notes to table 7.5 and de Vries, Europe in an age of crisis, table 1, p. 5. Over the full 200-year period the urban percentage quadrupled in England, scarcely changed in the rest of north-west Europe and advanced rather modestly in the continent as a whole. The English experience appears to be sui generis. The extent of the contrast also comes home forcibly if de Vries's estimates are reworked to permit another comparison to be made, as may be seen in table 7.7. The urban population of Europe more than doubled between 1600 and 1800, but the total population rose by almost 60 per cent so that much of the rise in urban population was caused by the rise in overall numbers rather than in increase in the proportion of the population living in towns. Column 4 shows how the urban population would have grown if the urban percentage had stayed at the level prevailing in 1600. The totals in column 5 show the 'net' gain in urban population at each later date compared with 1600, that is the number by which the urban population exceeded that which would have obtained if the urban percentage had not changed in the interim. The second panel of the table repeats the calculation for England. By combining information from the two upper panels, the proportion of total European urban growth that occurred in England can be calculated for the periods 1600-1700, 1700-50 and 1750-1800. The proportion rises steadily from 33 per cent in the seventeenth century to 70 per cent in the second half of the eighteenth. Over the two-century period as a whole the proportion exceeded one-half. Since England contained only 5.8 per cent of the population of Europe in 1600 and even in 1800 only 7.7 per cent, these proportions are an extraordinary testimony to the extent of the difference in urban growth between the island and the continent. De Vries laid particular emphasis on the absence of growth in the smaller towns in the period 1600-1750, which he defined as the age of the rural proletariat. He argued that urban growth was almost entirely confined to very large cities, with 80 per cent of all growth taking place in towns with 40000 inhabitants or more. These were almost all capital cities or large ports whose development was stimulated by the growth in administrative, military and legal employment in both absolutist and constitutional states, or by the Urban growth 178 1000 500 London 1800 London 1750 London 1700 CO —N. 0 00 N. 0)0) ('C') 000 '-o---ci 0 0 N. (0 U) OCIU) oo. 000- U) 0 Co NT (000)0) U)(DCO 0)0) 0) MC'J CO 0 U) C'J (\JCCL() 0000 London 1600 100 C 0C') London 1800 1700 1750 10 10 Rank Figure 7.2 Urbanization in England 50 100 development of long-distance, often extra-European trade. Smaller towns stagnated or lost population, afflicted by loss of political autonomy and by the 'abandonment of cities as locations for many of the most labour-absorbing industries' 29 The balance of advantage favoured protoindustrial development in the countryside. It is unquestionably true that there was much growth of industrial employment in rural England in this period, as we have seen, nor can there be any dissent from the view that London, combining administrative, commercial and trading dominance, enjoyed an astonishing expansion. 30 But neither the development of rural industry nor the growth of London precluded an equally remarkable surge of urban growth elsewhere in England. Other towns grew by much the same absolute amount as London between 1600 and 1750 aiid proportionally were growing even faster than London (table 7.2). Towns with between 5000 and 10000 inhabitants doubled in number from 15 to 31, while over the same period the number in Europe excluding England fell from 372 to 331.31 Figure 7.2 shows rank29 De Vries, 'Patterns of urbanization', p. 101. 30 See chapter 6. 31 De Vries, 'Patterns of urbanization', table 3.7, p. 93. For sources of data about English towns see source notes to table 7.1. The English share of European urban growth 1600-1800 (populations in millions 50 'IT — 0)0 CJ 0) CO N. c'j(000 U) Co N. LO (- LOcO U)NCO.- oo - ci C 0) C 0100)(O LO 0 (DC') N. 0000 -o It 0)0) 0 oqo,- CO 0600 ci) C C C') N. 0 C') (0C') U) N. U) Lo (q(CO 0 LU (0 N. N. Co '-ON-CO LU LO CO W '-0 '- C\J -'--00'N. 0)0) U) 0C')(COL( ..000o0000 (1) 0000 C 0000 0000 00U)0 00(00 (ON. N. CO D (ON. N. CO 0. 2 w C w 0000 0 U) 00 N. N. CO CO .00000 tO 00 U) 0 N. N. CD C w Urban growth Urban growth and agricultural change size plots for England. Like figure 7.1 it is schematically drawn: empirical data plots would show, of course, points falling on either side of the straight lines for each date. Although the number of towns is small and the presence of London severely distorts the top of the distribution, the lines emphasize how different England was from the continent (compare figure 7.1). Dr Vries's study has the admirable virtue of providing a framework of comparison for smaller units by epitomizing the characteristics of Europe as a whole. This helps to bring out the distinctiveness of English history, something which may be further examined by considering some comparative data about England's two greatest rivals, Holland and France. rural population grew by 49 per cent overall but rural agricultural population by only 13 per cent, even though the overall population growth in England during the century reached 71 per cent compared with a Dutch figure of only 11 per cent. If, therefore, it is safe to assume that Dutch rural agricultural population was rising as fast as rural population as a whole, a plausible supposition at a time of falling real wages, rural agricultural population in the two countries must have moved almost exactly in step in the eighteenth century even though their overall rates of population growth diverged so markedly. Conversely, in the sixteenth century while impoverishment increased in England and its rural agricultural population grew fast (by 58 per cent between 1520 and 1600), in Holland there can have been very little comparable increase, though the population of Holland was rising quite rapidly. In columns 5-7 of table 7.8, the rural population is subdivided between agricultural and other employment in confirmity with arbitrary assumptions whose basis is explained in the notes to the table. To the degree that the figures mirror reality they underline the points just made. The 'Golden Age' of the Dutch rural economy was clearly one free from increasing pressure on the land but it was succeeded by more trying times. 34 It would be absurd to press too far any comparison of the urban sectors in England and Holland. It would make no sense, for example, to try to use estimates of urban percentages in Holland as an indirect measure of agricultural productivity since she was a large importer of agricultural products, above all Baltic grain, and was not broadly self-sufficient as England was in the early modern period. Yet the beneficial effects of urban growth on Dutch agriculture in promoting specialization and making it easier to achieve higher production per man and per farm reflect the same processes at work in Holland as London's growth produced in England. Again, the Dutch passenger canal network was the transport wonder of its age,35 a response to urban growth and the closely associated rise in living standards, just as similar developments in eighteenth-century England promoted the investment of capital in turnpike roads and the construction of a new canal network. Both Holland and England, therefore, vividly illustrate the beneficial interaction between urban growth, rising living standards and a surge in agricultural productivity that was possible within the context of an early modern economy. The food needs of towns were met through the operation of a commercial market in foodstuffs by farming units which benefited from specialization and avoided the subdivision and fragmentation of holdings that was commonly the bane of peasant societies when population increased. By the late eighteenth century yields per acre were substantially higher in England and The Netherlands than elsewhere in Europe, having roughly doubled over the previous two centuries. 36 Yet in neither country was the 180 Holland and France Even in 1800 England was a less urbanized country than Holland. Already in the early sixteenth century many Dutch people were town dwellers and they increased in number very rapidly as the century progressed. In this period the course of real wages in Holland was almost the reverse of the comparable pattern in England: real wages rose to a peak about 1610, the date when the English series approached its nadir. Urbanization proceeded apace. The percentage of the population living in towns grew from about 21 to about 29 per cent, and since the total population of Holland was rising moderately quickly at the time, the absolute number of town dwellers rose very rapidly from about 260 000 to about 435 000. There followed a period of half a century during which Dutch real wages fell back somewhat just as a recovery was beginning in England, maintaining the inverse movement in the two countries for a full century. After the 1650s, however, Dutch real wages rose rapidly once more to a new, and substantially higher, peak in the 1690s, after which they tended to fall, uncertainly at first but more quickly and without interruption from the 1740s until the end of the century.32 The trend of urban percentages moved broadly in sympathy with real wage movements, reaching a high point about 1700, when about 39 per cent of the Dutch population comprised urban dwellers, but then slowly declining until by 1800 it was not much higher than the English figure. Some Dutch towns experienced very savage falls in population and, since the population as a whole grew only very modestly in the eighteenth century, the total number of town dwellers fell slightly.33 Dutch rural population trends also warrant notice. During the long period of urban growth from 1550 to 1700 rural population grew only modestly. It was 17 per cent larger at the latter than the former date. In the eighteenth century, however, when there was urban decline and real wages languished, rural population rose by 20 per cent. In England, in the eighteenth century, 32 Dc Vries, The pre-industrial Netherlands', pp. 671-4, and figure 2, P. 673. 33 See notes to table 7.8 for sources for these estimates. Van der Woude provides much detail of urban fortunes in Holland between 1525 and 1795 in 'Demografische ontwikkeling', especially pp. 134-9. 181 34 De Vries, 'The pre-industrial Netherlands', figure 2, p. 673, charts the fall in real wages from a late seventeenth-century peak which, in a period of static or falling population, is eloquent testimony to Dutch difficulties. 35 Dc Vries, 'Barges and capitalism'. 36 Slicher van Bath, Agrarian history, pp. 280-2. Urban growth and agricultural change WØ)'C CO C)C C.Ca5 .Q) n Ca) 2 Co a) E o 0 a)aa) (D 2E X-8 01 OaOa) LU (3) .-a)°o) 8 C IC) LO 0 LO LO LO 0 C) C\J NC'J Co qt (0) qt Nt - CJ CQ Q) 0)3(0 CoE il CO)D S Q .CC D C 8. -C- C2w o 3a) r.5?-ôE 5 E r.9E° C) U) LO 0 00 C'4 Cr) LO LU CO CO N- N- N- N- CO CI .12 CD ' 3 -.Oa) o>D E 2C .8-° 0 cC0 C) a) •0 C LU.- C'4a) C0>.O(O C0 ( LO LO 0 CO CO CO CO E 0,a)COOa) -o 2a)C (0€a)WE a) C220 8-C° r- r--: COCOCO (C) ci 00000 1: 0 0. 4) E o.8Z E CIL - . Da CO 10 a)CO 0 L 000 LO C) CO CO CO LU 0 Cl) 0 .0.52 C 0) 0 c C 2— C CO 03CO Q. .O Co a. HUflV I;!a) 2.2 IZ C,) a) E 0 LO Co ('J CO LU 0 LO LU C) - N- C) CO N- CO CO U, a) C 0 Ca a0 aCO C) N- C) LO CO -a -C 0 -o Co O CE ,2 co0— CO II E>, C 0. - CO, COa)0 8 O cE OD -C a)> D C0 D E 0) 2 a) .o .I 8 CO CO -C ci 00000 agriculturally employed population very much larger in 1800 than in 1600. The example of Holland also shows, however, that such progress is not necessarily a passport to further success. Adam Smith, aware like so many of his contemporaries of the exceptional economic achievements of Holland, used it as an example of the limits to growth which must beset any country at some stage, making it difficult to sustain the gains of the past, much less secure further advance.37 The tide began to ebb in eighteenth-century Holland. Urban population fell away slightly even though numbers rose modestly overall. Real wages sagged. Far from forming part of the vanguard of the industrial revolution, Holland moved into the industrial era later than most of western Europe, and even as late as 1850 was no more urbanized than she had been in the later seventeenth century. France makes a very different contrast to English experience. Some relevant data are given in table 7.9. The steps leading to the estimates given in the table are explained in the notes to the table. As with the English and Dutch estimates, they are subject to a margin of error which may in some cases be substantial because the empirical data base is insecure or inconsistent, because some of the assumptions are questionable or because the chains of reasoning used to produce some of the estimates were long and there may be a compounding of errors. Despite such uncertainties, the contrast between France and England stands out so strongly that no reasonable change in the assumptions used would significantly alter the picture. In 1500 France was not only a far more populous country than England but also more urbanized. Within the present borders of France there were then already 14 towns with a population of more than 20000 and 21 with populations between 10000 and 20000, while in England only London exceeded 20000, and probably only two other towns, Norwich and Bristol, had 10000 or more inhabitants. London was not only smaller than Paris but also smaller than Lyon.38 Overall, France was almost twice as urban as England, though much less urban than Holland, but in the next three centuries French towns grew very little quicker than the French population as a whole. France closely resembled de Vries's 'Europe' both in level of urbanization and in its change over time (table 7.5). By 1800 the degree of urbanization reached in England was approaching three times the French level.39 Moreover, the proportion of English rural population employed outside agriculture grew much faster than in France, and agricultural 0, 0, D O, C C C: .5? .Q -8 0 Ca C Ca E 183 E,,ot 2 8i COO) 8 00 LO 0 LO 0 LU 0 N- 0 CJ 0 C'J LO Co C) C) a - i .c tj 2co •o C° 0 3C -15 . CO 0, Cd E .2 8. a) C. a. 0 0, Q2 000000 LU0 Lo U-)0 Lo CO CO N- N- CO >, 0) o .oCo d .0 a) Ca 8. 0 .0 a) ' E 2 .2 .2 jE .; -00 0 CO ci t 37 Smith makes the general observation that, 'In a country which had acquired that full complement of riches which the nature of its soil and climate, and its situation with respect to other countries allowed it to acquire; which could, therefore, advance no further, and which was not going backwards, both the wages of labour and the profits of stock would probably be very low'. He subsequently suggested that Holland was approaching this state. Smith, Wealth of nations, i, p.106. 38 The French town population estimates are taken from de Vries, 'Patterns of urbanization', table 3.2, pp. 82-5. 39 It should be noted, however, that the effects of revolution and war caused a sharp drop in French city populations in the last decade of the eighteenth century in some cases. Dupâquier, La population francaise, pp. 91-2. (OCU)a) Co 5 C a)•_ 2a)oa) .2 n o- a) Q.ca0 a)1- c E S Co a)-a0 ' U)0 " 0D 0)E$ L o O)_ Ca)--a)C'JO on 0ô a) C o =N.5.0.Q 2a5D0...CL co a) 00000 C'J U) (N 0 CO (N U) 0) CO (N lqt U) Co CO -E 2! o.2?. -o 0 a) .2 o a) U).0 g)m Ca) a)a).O .0 U ) C2Ca).CDO D a) 0 N c cn a) a) a)D a)C. 0 .c • a) 2 a) C CL a)• ( C•x Z:.5 Z2! a) co OOO.. .0 c5 '-20 c.5 :2 C U) >2 >- c o0._a) C a) 2? 00000 N- 00 U) U) (Y) () 0)0 0 cu U) N- oo a) C.0° 00- CJ'-Co00 1 0 -CC e ":S ca aSa) CL 0) cc co 0) o U) in C)J C-.J . (0 C'Jt)JO) 0) U) CJ a) - . 5 . U o.0EU) oEEa) a) oa)a) E U) U)O Q) U) U) od000 2!a)E 0U).-ODCO CON- N- (0 Co a) . a)a)a)2o. a).5S-='-> 0 0)... CL . .0 a) 0 a)-5 c 0a) a)25 E A? U) U) -a)-0 a8m a).2a).a) 1g 00000 U) N- a) 0) 0C-0)Co qct N-(3) '- U) a.'-- -- c'Jc'J C .. a) 0 .5 CL -g 2 OC 0> 8 O02? oo .-uC .! 0 a)a) 200) - M 2? 2?Eo C c a g 2 -ac 2 3qj cc0) -M cc cc U) -0) IctU) U) N- U) cn c LL U -. U) - 0 C:J C\J CO C)J 0) IT 0) (0 N(00)'- 0 0) co (0 co COO).(0 (D cc C -- a) 0000 000 U) U) (0 N. N. cc E-a)5 .W fl a) o)a) 0 2 w0 a) )0 1 . 2 Ca 'a) -c'Jc'Jc'J D 5- o m -0g g DcC0C -CL a) ( D C .5 Ca)._.. a) U)C.5c\JU)2 CL Q).? U) a) C)C> 000 50?.00O -D a) M )a)a)0)cCcJ0a)00U) 00 a)a)a) 5I .0 0..Q)a) 0.. - QC )CL 2 2o LL 00 a)CQ) D)a) 0r LL -oC.0 a)'a)2? C r2!ra) CC CLc6 No5o DCDO - CC'J a)0a) lzr a) o C 0 0- (')CO. 00. .0a 3. a) .0 0. .cm - LL CL C CL E 3 - 2 LO 0' _3_co O.0 C C _1)C•C () (0 -a)a)0. .0 ELO Z3 w . .SCU)O. o.2 0 a S: 0. _a)Cu C a) 5 CT 2CJ a)-.-)C 0(0 a)a).C"52?a)0 a) > a) E2 "Wow• a) (I' U) a) > Q)O 2-2 -00 00000 000 U) 0 U) Co N- N- U) cc > .S2(0 (I 00000 00000 U) 0 U) U) U) 0).'- q 0) U) 5 C\JQ) 'C 0- E2.o ._a) • 0)(0 cQu acli a) 2?a C22 00000 Co U) C) (N Co () U) (N .-2 U)Q)a) cca)a)' > C.(0' a) of 0 U)a) C 0)0a)Cl) -ga)5 Lt)a) 0) a)U) Ec'i 0o C U- a)E cu a)a)CU) C.0.. I a)U)Ca)0U)0...0)(0 >0U) a)0 -- a)>Lt) 0 -0 .0 ã2 aa) ( LO S Table 7.9 (continued) French urban and rural population estimates (thousands) Ict 0) CO 00 Population totals by town size class (thousands) .00-(0a)a)0)CE 0 So -a)a)0-0 2 70 a).c 0)a) 0 a) 0 0 C5 8'U S 0 0 a) 15 CL -- . Ca)U) öa) .2.C)CC'J o -3a)€ 0 CL ca -N OC2?°O-Wa)..CO C 0 0. 0. Oa) rEa) '- -2 * g( : ° 2! ; Q.0•0u: ....5 o .0 .QOa) .0 0 .a . o0 02? 0 U) E.... co 0 .5< -U E V) '0 C a) -M C o 5 a).. 5 2Q> 0.0 a) LO I.T. co ?C C...a)(0a)00 N. 0 0 (0.0 Cl) n C') Urban growth Co C'J 0 CO C) CO CO 0.- C\J U) N- 0.- CO CO COO CO COO C'J ) CO CO 00.-.-C'.J (00 U) 0 CO 00.- CO ICT U) 0 ('.1 C\1 CO 0 U) C) It 0 CO CO CO CO 00 C') CO CO 00000 000 U) 0 U) CO N- N- CO CO U) CO CO 0 CO qCT U) CO N- Co CO 0 U) C') 0.U) 0 C'.J C-\J ICT.— C\J CN CO C\J C) CO CO CO .— Uj 00 CO C) C'.J C) 0 - U).- U) CO 0 U) N- C"J C) CO 00 C) C) 0 CO 000 Cr) 0 C') 00 U) U) 0 C'J C'J CO N- 00000.C'J 0 N- 0 U) 0 CO CO N- N- CO U) population per 100 agricultural productivity, using the very rough measure employed faute de mieux in this article, progressed far more rapidly in England than in France. Population always grew faster in England than in France, at times dramatically so (tables 7.2 and 7.9), yet because of the far faster urban growth in England and the swifter rise in the proportion of the rural population employed outside agriculture, the growth of rural agricultural employment was actually slower in England than in France apart from the sixteenth century. Between 1600 and 1800 the number of those dependent upon agriculture for a living in France grew by 30 per cent, whereas in England the comparable figure is 9 per cent (over the same period the national population growth rates were 53 and 111 per cent respectively). These apparently exact percentages should be viewed with reserve, but it is probably at least safe to assert that the rural agricultural population in England grew no faster than its counterpart in France. If the estimates given are taken to be broadly trustworthy, inferences about comparative trends in agricultural productivity in the two countries are possible. The upper panel of table 7.10 sets out information for England and France in an indexed form to make it easier to appreciate the divergent course of development that characterized them. The value for 1600 has in each case been made equal to 100. In the sixteenth century the differences between the two countries were chiefly related to the much higher rate of growth of the English population rather than to differences in the relative proportions of the populations falling into the urban, rural agricultural and rural nonagricultural categories. After 1600 the contrasts between the two countries grew more pronounced. England continued to grow the quicker overall but the rural agricultural element in her population grew very little over the next two centuries. Indeed, it appears to have been falling for much of the time, whereas urban growth was meteoric and the rise in the rural nonagricultural component was also pronounced. Growth in France was much more balanced, and as a result, although the total population grew less than in England, the rural agricultural population grew appreciably faster. If it were safe to assume that the level of food consumption per head was much the same in the two countries, that the estimates of rural agricultural population are accurate, and that each country may be regarded as supplying its own food needs (except that in table 7.10 imports were assumed to meet 10 per cent of English food requirements in 1800), then it is a simple matter to compare both levels of productivity in agriculture and their rates of change over time in the two countries. They are given in the lower panel of the table. The figures in columns 1 and 2 represent an absolute measure of productivity per head of the rural agricultural population while those in columns 3 and 4 index changes over time. The extent of the contrast between England and France may well be understated by table 7. 10, both because it is reasonable to suppose that consumption of food per head was somewhat higher in England than in France since English real incomes appear to have been the higher of the two, and also because a higher proportion of the output of English agriculture English and French growth patterns compared 186 0000.000 U) 0 CO N- N- CO U) Urban growth Urban growth and agricultural change may have been used as industrial raw material rather than food. Land and labour used for the raising of sheep for wool, for example, cannot also be used to grow corn. The effect, however, is difficult to quantify since output in such cases often consisted both of food and of industrial raw materials. Producing wool for cloth or hides for the leather industry also means producing meat for human consumption. 'Pure' industrial land usage, as with flax for linen, is a rarer phenomenon.40 The use of the land to provide 'fuel' for horses or oxen also suggests that the contrast between the two countries may be greater than appears in table 7.10 since in England by the later eighteenth century a very substantial proportion of horses were employed outside agriculture altogether on turnpike roads, on canal towpaths, but particularly in urban transport of men and goods. 41 In England, as in Holland at an earlier date, urban growth, rising productivity in agriculture and improving real incomes were interwoven with one another, and were mutually reinforcing for long periods. The French economy was not drawn into the same pattern. On the assumptions used in constructing table 7.9, there is evidence of a significantly increased pressure on the land in the eighteenth century. In a peasant economy this will tend to provoke serious economic difficulties and increase the proportion of excessively subdivided holdings, conforming to de Vries's 'peasant' model of a rural economy while both England and Holland accord better with his 'specialization' model •42 Arguments couched in this manner are always too simpliste. It is misleading, for example, to treat pre-industrial countries as if they were homogeneous units. There were major differences between different areas even within the smallest of the three countries considered. Sometimes even intra-provincial differences in Holland were marked: inter-provincial contrasts could rival those between countries. 43 The regional variety of a country as large as France was still more notable. It could scarcely be otherwise when it is recalled that France contained more than ten times as many people as Holland and that her area was fifteen times greater. 44 Furthermore, better knowledge would no doubt cause significant alterations to be made to many of the tables on whose contents the arguments advanced in this essay were based. Yet the contrasts and similarities between the three countries would probably remain even if perfect knowledge were miraculously to supervene. 188 40 Thirsk presents good reasons to suppose that the value of industrial crops rose very rapidjy in the later sixteenth and seventeenth centuries, and also provides abundant evidence of the enormous increase in rural employment outside agriculture. Thirsk, Economic policy and projects. She also makes some guarded attempts to quantify the scale of industrial crops and 'new' food crops. Ibid., pp. 177-8. 41 Thompson estimates that there were almost half a million (487 000) horses in use outside agriculture in 1811 (riding horses, carriage horses, post horses, trade horses and those in the stage coach trade). At the level of consumption of oats and hay per animal which Thompson thinks appropriate, and on reasonable assumptions about yields of oats and hay per acre, sustaining such a large horse population must have meant devoting about two million acres to their fodder. Agriculture at that date is estimated to have required about 800 000 horses. Thompson, 'Nineteenth-century horse sense', p. 78 and table 2, p. 80. 42 Dc Vries, The Dutch rural economy, chapter 1. 43 The work of the Wageningen school illustrates the point: Sli&er van Bath, van der Woude, Roessingh, Faber. 44 For population totals see tables 7.8 and 7.9. The area of France today is 212 209 square miles: that of Holland 13 967. Their areas varied somewhat, of course, over the early modern period. 189 Conclusion It is not difficult to find convincing reasons why a major advance in agricultural productivity is a prerequisite of industrial growth. Only if resources can be spared from the task of ensuring an adequate supply of foodstuffs can a larger scale of industrial production be attempted. It may seem quaint that Adam Smith should have associated the prosperity of, say, Sheffield's industry with the efficiency of agriculture in the vicinity of the town, but his remark exemplifies the circumstances needed for industrial success in a pre-industrial age.45 England was singularly fortunate in this regard. Output per head in agriculture appears to have risen by at least three-quarters between the beginning of the seventeenth and the end of the eighteenth century. The annual gain in agricultural productivity per head was modest, about 0.3 per cent, yet it permitted a more rapid overall rate of population growth than that found in either France or Holland while simultaneously releasing into secondary and tertiary employment a far higher proportion of the active population than in any other European country. By 1800 little more than a third of the English labour force was engaged in agricultuie at a time when it is improbable that the comparable figure elsewhere in Europe apart from The Netherlands was less than 55-60 per cent. In many countries it was substantially higher.46 The pace of English population growth in the early modern period was quite exceptional. Between the mid-sixteenth and the early nineteenth century her population grew by about 280 per cent. The population of other major European countries rose much less rapidly. Germany, France, The Netherlands, Spain and Italy all grew by between 50 and 80 per cent over the same period.47 The demographic mechanisms by which the growth of England came about are now fairly well understood, but at one remove the 45 See note 26 above. 46 In Finland in 1754, 79 per cent of all economically active heads of family were engaged in agriculture, forestry or fishing: in 1805, 82 per cent. There are data for many European countries from about the middle of the nineteenth century. In the following list the percentages refer to the proportion of the total labour force (male and female) engaged in agriculture, forestry or fishing unless otherwise indicated: Belgium 1846, 51 per cent; Denmark 1850, 49 per cent (family heads); France 1856, 52 per cent; Ireland 1851, 48 per cent; The Netherlands 1849, 44 per cent; Spain 1860, 70 per cent (males only); Sweden 1860, 64 per cent; Great Britain 1851, 22 per cent. Mitchell, European historical statistics, table Cl. Half a century earlier it is virtually certain that each of the comparable percentages would have been higher, sometimes substantially so. 47 See below, pp. 215-19. Urban growth Urban growth and agricultural change problem of explaining the contrast remains.48 In this connection it is important to pay heed to the probability that there was nothing unusual in the rate of growth of English rural agricultural population, unless it was that it grew so slowly, especially after 1600. The great bulk of the overall increase took place in that part of the population which made its living outside agriculture. This was what made England so distinctive. In one sense an explanation of English agricultural improvement may be sought and perhaps found in the relatively rapid and uninterrupted growth in the urban sector of the English economy, well able to afford food but producing none itself. The extraordinary stimulus to development afforded by the growth of London, which in 1500 was not even one of the dozen largest cities in Europe but by 1700 was larger than any other, was probably the most important single factor in engendering agricultural improvement. The momentum first given by the growth of London was carried forward later by the more general urban growth of the later seventeenth and eighteenth centuries. It encouraged both change within agriculture sensu stricto and in a host of associated institutions and activities: transport improvement, credit facilities, capital markets, commercial exchange. The joint effect of rising productivity and increasing urbathzation also fostered the gi#at expansion in rural employment outside agriculture. The combination of a steadily rising volume of demand for goods and services other than food with a far more sophisticated market mechanism for exciting and satisfying such a demand was the basis of prosperity for the industries in the countryside. Yet the reasons for the exceptional growth of London from the early sixteenth to the late seventeenth century remain imperfectly understood. Its rise both deserves and demands much more attention than it has received. Understanding of the phenomenon is likely to benefit especially from study of the connection between the development of the nation state under the Tudors and Stuarts and the growth of the capital city, and from considering London's fortunes within a wider framework of analysis of European commercial exchange, but the bulk of the task remains to be attempted. Urban growth towards the end of the early modern period presents problems of interpretation no less difficult than those at its beginning. During the eighteenth century the urban hierarchy of England was turned upside down. The new industrial towns and port cities of the north and the midlands thrust their way past all rivals other than London. Only Bristol and Newcastle among the traditional regional centres matched the new challenge, and then only because they could benefit from the stimuli that had forced the pace of the growth in places like Liverpool and Sunderland. Many once-great centres were on the way to the pleasant obscurity of county rather than national fame: York, Exeter, Chester, Worcester, Salisbury. Certainly the upsetting of the old urban hierarchy in England was at the time an event without recent precedent in European history. Elsewhere the exact ranking of major cities in each country varied somewhat from time to time but it was very rare for tiny settlements to develop into major centres. The same lists of large urban centres may be found century after century slightly rearranged. Occasionally, one of the smaller centres made progress through the ranks. Atlantic and colonial trade brought Bordeaux and Nantes rapid advancement between 1500 and 1800, for example, but revolutionary changes were rare. The progress of the new centres in England was such, however, that not merely had Liverpool and Manchester outpaced all their English rivals other than London in 1800, but by 1850 they were the seventh and ninth largest cities in Europe, and the largest anywhere in Europe other than those that were capital cities,49 an extraordinary tribute to their economic vitality unassisted by the employment in government, the professions and the arts associated with capital cities. It is evident that the history of urban growth in England was quite distinctive. The contrast with the course of events elsewhere in Europe was especially notable between 1600 and 1750 since the paralysis which appears to have affected all but the largest towns on the continent was absent in England. The expansion of secondary employment in the countryside was not a bar to urban growth in England in the manner hypothesized by de Vries for Europe as1a whole. In this there may be an important clue to the differences between England and her continental neighbours in real income levels or trends, in relative wage levels in town and country, in the terms of trade between the two, in urban institutions or in still other factors. In this essay, however, I have been primarily concerned with the narrower topic of the scale and speed of urban growth and the indirect measurement of the gain in agricultural productivity per head, which probably roughly doubled in England between 1600 and 1800. By the standards of the recent past a doubling in agricultural productivity over a two-century span of time is not a startling achievement, but in terms of the pre-industrial world it is much more impressive and set England apart from the great bulk of continental Europe, conferring on her advantages denied to economies unable to release more than a tiny fraction of the labour force from work with the flockor plough.50 Measuring agricultural advance through urban growth, of course, falls well short of explaining either phenomenon. It is easy to see how these developments might reinforce one another, but the fact that similar cases were so rare, and that when they did occur they tended to lose momentum after a while, suggests that noting the logical possibility of such a link is only a first step towards understanding what took place. Perhaps the most 190 48 The secular changes in fertility and mortality which largely governed population growth rates in England are described in Wrigley and Schofield, Population history of England, chapter 7. 191 49 Chandler and Fox, Urban growth, p. 20. The first nine in order of size were London, Paris, Constantinople, Saint Petersburg, Berlin, Vienna, Liverpool, Naples and Manchester. 50 Grigg provides information about rates of growth in labour productivity in agriculture in Denmark, France, the United States and the United Kingdom in the late nineteenth and early twentieth centuries, and for all major subdivisions of the world since 1960. Grigg, Dynamics of agricultural change, tables 21 and 22, pp. 171-2. Rates as high as 5 per cent per annum have been commonplace in developed countries in recent years. Urban growth Urban growth and agricultural change fundamental and intriguing question concerns the circumstances in which Ricardo's law of declining marginal returns to additional unit inputs of labour and capital may be circumvented in a land long fully settled. This law, if it had universal and invariable application, would prohibit a sustained development of the type which took place in England. The population of England increased more than three- and- a-half times between the early sixteenth and the early nineteenth century without becoming dependent on food imports other than marginally. Some new land was taken into farming use, but the bulk of the increased output must have been obtained from land already in cultivation. Moreover, after 1600 there was little increase in the agricultural labour force. What served to neutralize the operation of declining marginal returns? The usual answer to this question is to point to innovation. If new methods of farming were introduced which significantly enhanced the productivity of labour and capital, then the operation of Ricardo's principle can be postponed, indefinitely if the flow of suitable innovations is sufficiently sustained. In a sense this must be the right answer but because it is inescapable it may be unilluminating, for though it may be logically necessary it was historically contingent. Relief from such a source was not always or even commonly forthcoming. Can anything be found in the circumstances of early modern England which provides an answer to the conundrum? To attempt a full examination of this issue is beyond the scope of this essay, but one of its aspects may bear a slight elaboration by way of conclusion. There was a remarkably strong and regular positive relationship between the long-term rate of growth of population and the comparable rate of change in food prices in England throughout the early modern period down to 1800.' At first sight this suggests a uniform and inhibiting tension between population growth and the capacity to sustain rising numbers. But, although the tension was constant, it does not follow that all the demographic and economic variables involved stood in the same relationship to each other throughout. We have seen, for example, that there were major changes in national occupational structure and in agricultural productivity per head. The relationship between population growth and food prices may prove to be common to England and other countries, but the changing structure of the component variables was certainly not usually found elsewhere. The tension appears to have been beneficial and dynamic in England, when so often it was debilitating and severely inhibited growth. Innovation may be the key to overcoming Ricardian constraints, but inasmuch as what happened in England was so rarely paralleled in other pre-industrial economies, it seems doubtful whether innovation could be counted upon to be forthcoming in response to need. When, as in England, events took a more favourable turn, it may be that the explanation is not to be found in the urgency of human need, nor in the immediate price dynamics of the market place, nor in the accident of individual inventiveness, but in the unusual structural characteristics of the prevailing situation. If it were safe to assume that a community always exploited its knowledge of production methods to the full, it would follow that innovation meant the introduction of some element into the productive system which was previously unknown. On the other hand, if there are a range of other options capable of raising productivity which are held in reserve, so to speak, because current circumstances do not encourage or even allow of their use, the nature of the case is cianged.52 It may not be sufficient to look exclusively at changes in technology to explain a rise in productivity: new crops, altered cultivational practices, improved breeding techniques and so on; nor to turn to the gains in productivity per head arising from increasing specialization of function, the chief proximate cause of enhanced productivity in Adam Smith's analysis of the question. Other developments need to be taken into account if they encourage the exploitation of previously untapped potential and thereby transform living standards and growth prospects. It is in this context that the exceptional scale and speed of urban growth and its particular nature were probably so important.53 192 51 Wrigley and Schofield, Population history of England, figure 10.2. The relationship between the rates of change of the two variables from the mid-sixteenth to the late eighteenth century was strikingly linear. This in itself may be suggestive. If there had been very tight constraints upon agricultural expansion, a curvilinear relationship might have been expected with a steeper rise in price than in population as conditions worsened. Its absence may reflect the existence of an unusual capacity to rise to the challenge of increased demand on the part of English agriculture, especially when it is remembered that the absolute rate of population growth was at times quite high, approaching I per cent per annum in the late sixteenth century and exceeding it in the late eighteenth. I owe this comment to Dr Schofield's reflections on the point. 193 52 Some years ago Boserup made very effective use of one form of this argument, but her 'model' was one in which the changed methods of agriculture adopted to meet the challenge of rising population resulted in the broad maintenance of a given level of output per head per annum at the cost of a rise in the number of hours worked per annum. I have in mind changes which substantially increase labour productivity whether measured by the hour or by the year, though a part of the latter may be achieved because the workload becomes less markedly seasonal and therefore there are fewer periods during the year when labour is intermittent. Boserup, Conditions of agricultural growth, especially pp. 41-55. 53 Some further aspects of the gain in agricultural output per man and per acre are examined i...