|
|
Dear All,
I have deeply enjoyed the discussions so far, and applaud the efforts of my
friends and colleagues to bring light to this difficult issue. Let me also
dedicate my comments to my dear friend Eric Jones, whose expertise on the
"process" of agricultural improvement in England has brought so much to light.
Where are we in this debate so far?
Clark claims there was no agricultural "revolution," just a slow
but steady increase in productivity and output reaching back into the
seventeenth or possibly sixteenth centuries. Turner sort of agrees,
suggesting "agricultural transition" as the right term to describe notable,
but not revolutionary, gains in output and productivity. Allen and Overton
insist there was at least one "agricultural revolution," consisting of
dramatic gains in output and productivity, and a sharp reduction in labor
force in agriculture. Allen adds a second "landlord AR," consisting of
shifts in landownership, with few changes in productivity or output, in the
late eighteenth century and associated with Parliamentary enclosure. Brunt
argues that there is insufficient data to settle this matter, although
noting a number of technological innovations and old techniques (e.g.
liming and marling) that would have undoubtedly raised output when applied.
Dare I say it--all these views are true! How can that be? The oldest
story in the book: things changed at different rates and in different ways
depending on where you look. But I say this not to discredit or obscure a
careful analysis of change. Rather, such regional variation is the key to
understanding the overall evolution of agricultural output -- AND ITS
REVOLUTION -- in England 1500-1850.
Let me try to make the following points: (1) There was indeed an
agricultural revolution in England, dating mainly to the period 1600-1750.
(2) This revolution is not visible mainly as an increase in grain yields
per acre or per man-years of labor. Rather, it is visible mainly in
conversions of land from pasture to arable (and to a lesser degree, from
arable to pasture), and in consequence, in increases in total output that
cannot be measured simply in terms of productivity or yield.
(3) This revolution was concentrated in the upland, chalk and sandy soil
regions of England. However, it took place against a background of steady
gains in productivity in the valley and clay regions as well.
(4) This revolution took place mainly in "old-enclosed regions," rather
than in old open-field areas. While this change was not due to enclosure
per se, it helps to explain the misleading impression held by
contemporaries that enclosure was associated with remarkable gains in
productivity.
(5) Despite many comments to this effect, the agricultural revolution did
not "release" labor for urban and industrial growth. Rather, the
population working in agriculture continued to grow throughout this entire
period.
What did occur is that the proportion of England's total population working
in agriculture was reduced, but that is not at all the same thing as a
reduction in agricultural labor.
(6) When the whole picture is seen, we can explain not only the errors
about enclosure, but also why England and not France had this kind of
agricultural revolution in this period.
(7) The whole debate needs to pay much more attention to the work of Ann
Kussmaul, whose work on changes in regional farming patterns in England
should be central to our understanding of this issue.
I will follow Liam Brunt in relying on some simple hypothetical
paired comparisons, that I hope will put micro-flesh on the macro-bones of
the findings reported by Allen and Overton and Clark.
To begin, let me ask your indulgence in letting me quote from a
ten-year old article on this subject in an obscure journal: "Regional
Ecology and Agrarian Development in England and France" (_Politics &
Society_ 16 [1988]: 287-334). Comparing the extent of arable (mainly grain
producing) and pasture land in England, I noted that "around 1600 arable
lands were concentrated in the Midlands, with extensions into eastern
Yorkshire and northern Norfolk. But a century later, the amount of arable
land had greatly expanded, particularly in the West Riding of Yorkshire,
Cambridgeshire and Lincolnshire, Essex, the Home Counties, Devon and
Cornwall, and the western Midlands." Thus much land was converted from
pasture to arable during the seventeenth century. Moreover, looking at
maps of grain yields and Parliamentary enclosures shows "virtually no
correlation between yields and enclosure." Thus to understand the
agricultural revolution, the key is not to focus on enclosure or even on
changes in yields, but rather to understand the causes and consequences of
the pasture-to-arable conversions in the seventeenth century, which were
quite significant and are striking on any map of English land-use.(see
Kussmaul and even Thirsk's agrarian history of Eng & Wales for excellent
maps of these changes).
Let us now play with a few hypothetical examples of what might have
lay behind these shifts. Let us compare a 15 acre swatch of land (it might
be one farm, or a small part of a larger farm) in two different regions c.
1600. Region 1 is a midlands, open-field manor farm, on high-quality clay
soils. It has long been cleared and worked for commercial grain
production, with some animal production as well. Region 2 is from an
upland, or chalk, or sandy soil region. It is farther from markets, and
much of its soil is too poor for grain production, and fit only for rough
pasture. Most of the grain is produced for local consumption, though some
animal products are sold for market. Here are some purely figurative (my
apologies to Mark, Bob, Greg, and others who have worked with more exact
numbers from actual farms recently) numbers.
Region 1, c. 1600:
10 acres in wheat/barley produce 16 bushels/acre, using 2 man-yrs. labor
5 acres in meadow produce 1 unit animal products (AP), using 0.25 m-y
labor Total Output: 160 bushels grain plus 1 unit AP Productivity figures:
16 bushels/acre; 80 bushels/m-y
0.2 units AP/acre; 4 units AP/m-y
By comparison, in Region 2, c. 1600:
5 acres in wheat/barley produce 12 bu/acre, using 1 man-yr labor
(half the acreage uses half the labor, but poorer soils have less
yield) 10 acres in rough grazing produces 0.5 units AP, using 0.25 m-y labor.
(twice the acreage but not tended meadow produces much less
fattening for animals, but requires about as much labor to watch and
process the animals).
Total Output: 60 bushels grain plus 0.5 units AP Productivity figures:
12 bushels/acre; 60 bushels/m-y
0.05 units AP/acre; 2 units AP/m-y
Now, set the clock in motion. The midlands region benefits from continued
efforts to improve the soils and seeds, and continues to market its grain.
As Greg Clark suggests, there is slow but non-revolutionary progress in
productivity and output, such that by 1750, the land's output has risen 50%
(not much of an annual rise, indeed an imperceptible .03% per annum over
150 years). In addition, as Greg also points out, productivity and returns
on meadow change hardly at all. We then have for
Region 1, c. 1750
10 acres in wheat/barley produce 24 bushels/acre, using 2 man-yrs. labor
5 acres in meadow produce 1 unit animal products (AP), using 0.25 m-y
labor Total Output: 240 bushels grain plus 1 unit AP Productivity figures:
24 bushes/acre; 120 bushels/m-y
0.2 units AP/acre; 4 units AP/m-y
This is nice, but certainly not an Ag Revolution.
But let us note that farmers in region 2 have a different choice. They
know that it is possible to convert their rough pasture to arable, by
growing green manure and fodder (e.g. clover, sainfoin, turnips),
stall-feeding their animals and concentrating the latters' manure on
fields, and adopting more complex crop rotations to nitrogren-enrich the
soil. Such a change requires a large investment, however, and is not
worthwhile unless they have better markets for the grain they would then
produce. Fortunately, over the course of the sixteenth and seventeenth
centuries, the London market (and lesser markets in Bristol, Norwich, York,
and other urban centers) extends its tentacles throughout England. Farmers
in the Home Counties, Cambridgeshire and Lincolnshire, even the West
Midlands and Devon and Cornwall and Yorkshire, increasingly find it
worthwhile to make the investments required. This of course is George
Grantham's argument, and I endorse it whole-heartedly -- the technology
here is neither newly discovered nor suddenly diffused; rather it follows
the growth and stimulus of the market. We then might have for a farm in
Region 2, c. 1750
12 acres in wheat/barley produce 18 bushels/acre, using 2 man-yrs. labor
3 acres in fodder crops produce 0.5 units AP using 0.5 m-y labor Total
Output: 216 bushels grain plus 0.5 units AP Productivity figures:
18 bushes/acre; 108 bushels/m-y
0.167 units AP/acre; 1 units AP/m-y
This farmer has converted most of his land to grains, relying on
stall-feeding of fodder crops to sustain his animal output, which he can
now do with about one-third the land of untilled rough pasture, albeit with
twice the labor.
Examine the consequences for productivity figures. Overall, looking only
at wheat productivity, this has improved in region 2 from 12 bu/acre to 18
bu/acre, or no more than in region 1. Looking only at grain productivity
figures, one would see no difference here from what was going on in region
1. With regard to labor productivity, things have improved much more
dramatically, nearly doubling from 60 bu/m-y to 108 bu/m-y. But this is
still below the labor productivity of region 1's 120 bu/m-y. And this
doubling of labor productivity for these lands would be obscured in
aggregate accounts by the 50% increase in labor productivity in the larger
base of region 1. Thus at 1600, total grain output for both farms is 220
bushels produced by 3 m-y, or 73.33 bu/m-y. At 1750, total grain output
for both farms is 456 bushels produced by 4 m-y, or 114 bu/m-y, for a total
increase in labor productivity of 55.5%
But looking at output tells a rather different story. The farm in region 2
has nearly QUADRUPLED its grain output, from 60 bushels to 216 bushels.
Taking our two farms together, even though grain output per acre has grown
only 50% and grain output per labor unit has grown by 55%, total grain
output has more than doubled, while animal output has remained the same.
Of course, while I claim these figures are plausible, they would not
necessarily be typical. In periods when animal prices were rising faster
than corn prices, farmers in region 2 would likely keep far more of their
land in pasture, but now in fodder instead of rough grazing, to boost their
animal output. Doing so has the by-effect of improving the land long-term
for future arable farming. When grain prices start to rise, conversions
such as the one I noted above are more likely. The precise mix of what
happens will vary from farm to farm and time to time, depending on local
and national market conditions.
Is this sketch at all plausible? We are unlikely to know from data on
individual farms. The best data, after all, tend to come from
long-established farming areas, where the land was held by manors,
charities, or ecclesiastics, and this was mainly land from region 1.
Region 2 land, in contrast, was more marginal, likely to be held by yeomen
and small-to-medium farmers. In fact, the tendency is for Region 1 land to
be mainly open-field with strong manorial control, while Region 2 land was
settled piecemeal, through enclosures of cleared forest or new settlement,
and was likely to show enclosure from the early middle ages onwards. This
made it easier for individual farmers to take the decision to shift their
crop rotation and land use if market conditions made it desirable to do so;
but the enclosure was generally already there, and not the causative factor
in the land-use or output change.
Still, if it WAS the case that the old-enclosure areas had enterprising
farmers who from roughly 1500 to 1650 could roughly double their output in
a short period of time due to radical changes in land use, while old
open-field farmers were making only gradual and imperceptible progress in
yields, then this could readily have created a misleading impression among
contemporaries that enclosures were associated with dramatic progress in
farming as opposed to "stagnation" in the open-fields.
But to return to the question -- is something like the above at all
plausible? The best answer is to quote keen observers, such as Daniel
Defoe, who could scarcely contain his astonishment at this change. Writing
of the downs of Hampshire and Wiltshire, he noted that they produced
"excellent wheat, and great crop too, tho' otherwise poor barren land, and
never known to our ancestors to be capable of any such thing; nay, they
would have laughed at anyone that would have gone about to plough up the
wilde downs and hills, where the sheep were wont to go." (the quote is
taken from C.G.A. Clay, Economic Expansion and Social Change, England
1500-1700, volume 1, p. 111.) Andrew Yarrington, an 18th-century
agronomist, was among those who emphasized "above all, the increase in
tillage which the use of clover made possible." In the Oxfordshire uplands,
the proportion of land sown to wheat doubled in the late sixteenth and
early seventeenth centuries, and in Hertfordshire, the eighteenth century
agronomist J.H. Smith (quoted by Eric Jones) noted: "the dry grounds, which
formerly were lett for a trifling rent, are now lett at twenty shillings
per acre; since the introduction of clover and turnips into their poor and
barren hills, while the low lying stiff grounds [i.e., clays], pay only ten
shillings, which is the rent they gave near a century ago." And in
Gloucestershire, "thousands and thousands of acres of pitiful, poor, sorry
ground not worth the ploughing and sowing with corn alone .. being enclosed
and sown with some of the seeds [of sainfoin, clover, and ryegrass] brings
forth so great an increase that the same land which before was hardly worth
ten groats an acre will not not be let [but] at three times ten shillings
an acre."
This helps resolve the Clark paradox-- yes from 1600 to 1750 or
even 1800 yields per acre and per laborer went up only 50% or so, and
usually imperceptibly on most lands. BUT, the conversion of much past
"poor, sorry ground, not worth the ploughing" to productive grain land
meant that gross output could easily double. I would accept that for many
workers, living conditions did not improve up to 1820, and that imports
greatly increased, so that gains in English output cannot be equated with
increases in population. Still, the capacity for English agriculture to
feed far more people extended well beyond what can be calculated from
changes in yields per acre or per laborer.
Now, to tie up some loose ends:
If the sketches above look more plausible, let us look once more at the
figures for animal production. Note that even c. 1750, animal productivity
of Region 1 remained higher than in Region 2. To review the productivity
figures c. 1750:
For grain, region 1's output per acre is 33% higher than that of region 2,
and output per man-year is about the same. But for animal products, region
1's output per acre is 20% higher, but -- and this is critical -- output
per man-year is double that in region 2. The reason is simply that fine
meadow land is more efficient for fattening animals than fodder
stall-feeding, in terms of labor required to till and move the feed. If
this was so, that the productivity of meadow land remained much greater in
region 1 while converging with respect to grain, then you would expect
landlords in region 1, from the mid-eighteenth century or whenever this
shift became apparent, to convert their land from arable to pasture. And
this is precisely what they did. This of course does not change overall
productivity figures (pace Allen), but it does give them a better return,
and motivate Parliamentary enclosure to force the arable farmers off their
lands.
Thus, again to quote contemporary observers: Pitt noted in 1809 that "the
course of agriculture [in the Vale of Belvoir in Leicestershire] has since
the enclosure turned topsy-turvy, the richest land in the vale, formerly
tillage, has been laid to grass, and the skirtings of the vale, formerly
sheep-walk, have been brought into tillage." (I owe this quotation to M.
Turner's book on Enclosure).
This change was only possible, following Grantham, because so much of the
English light-soil regions came within the pull of the London (and other
urban) markets from 1600 onward. In France, by contrast, the region 1
lands were concentrated in reach of the Paris market. These areas of the
Ile de France, as Phil Hoffman has shown, both had reasonably high yields
(comparable to midland England), and almost imperceptible increases in
productivity. Much the same was true of Normandy, the other main area in
reach of the Paris market. The areas of France that specialized in
livestock but that could have increased output by converting to grain, the
bocages of the west and center, were simply too far or too arduous a
transportation trek to France to repay investments in increasing output for
market. When the railways changed this fact, French productivity responded
in dramatic fashion. But before 1850, the logic of the Agricultural
Revolution in England shows why there could be no similar AR in France --
the combination of poor soils and moderate yields with great potential for
improvement PLUS easy access and pull of a major urban market simply were
absent in the latter country.
Finally, the notion of "releasing" labor from agriculture is not quite right.
If we accept Allen's and Overton's estimates of the proportion of the
population in agriculture as:
c. 1500 74%
c. 1750 45%
c. 1800 30%
and we multiply these figures by population estimates for these years 2.3
million, 5.7 million, 8.6 million (Wrigley/Schofield), then we have the
estimated population engaged in agriculture as:
c. 1500 1.7 m
c. 1750 2.6 m
c. 1800 2.6 m
In other words, during the height of the agricultural revolution (which I
argue was real) from 1500-1750, the number of agricultural workers
increased by over 50%. [Thus in my hypothetical example above, the total
labor employed by the two farms went from 3.5 m-y to 4.75 m-y, actually a
bit low]. From 1750 to 1800, the labor actually "released" from
agriculture, to the extent the estimates are accurate, was virtually nil,
in a population that increased by almost 3 millions.
These figures may in fact be rather high; E. Evans gives estimates for
England's labor force in agriculture of 1.7 million for 1800 and 2.1
million for 1850.
But the point remains the same--England's urbanization and
industrialization did NOT come from any marked reduction of toilers on its
lands. Rather, what happened was (a) the non-agricultural population
reproduced itself through rapid growth rates (viz. David Levine) and (b)
most of the additional children of agricultural workers were no longer
absorbed on the farms. The link between the agricultural revolution and
industrial revolution is thus much more complex than simply a "release" of
labor.
Industrialization created both a demand for more urban labor and urban
markets that demanded (and had more products to give in exchange for) more
food. As farms adapted to the greater demand by raising their output while
only modestly increasing their labor use, their greater productivity
facilitated urban growth. Because England had so much "poor" land capable
of improvement in close proximity to growing markets, given the river and
coastal access of most of the country to London, England's response to such
stimulus was more extensive than in any other large European country.
The higher agricultural output that resulted then allowed England's growing
population to not only continue rapid growth but to move to urban areas and
to non-agricultural pursuits. This was a virtuous circle of stimulus and
response, rather than a simple "push" from one direction or the other.
Best to all,
Jack Goldstone
UC-Davis
|
|
|
 Send comments and questions to admin@eh.net
|
||