Professor John Williams’s essay ‘The move from the land’ was ﬁrst published in Wales 1880–1914 (1988), a volume of the ‘Welsh History and its Sources’ series. It is not surprising that it was Williams (1927–2004) that the editors invited to write an essay on population movements in Wales in the nineteenth and early twentieth centuries for their volume. For decades until his death he was the foremost authority on modern Welsh economic history.
Born and raised in Cardiff, Williams was Professor of Economic History at Aberystwyth University. He published extensively on various aspects of Welsh economic, industrial, labour and social history. His books include The South Wales Coal Industry 1841-1875 (1981), co-written with J.H. Morris. He also published widely on British and international economic and social history but always retained an interest in and love for the history of his own country. Several of his important and thought-provoking essays on Welsh historical topics were reprinted in Was Wales Industrialised? Essays in Modern Welsh History (1995). His massive contribution to the study of the history of Wales lies not only in his publications but also in his tireless work for Llafur: The Welsh People’s History Society, of which he was Chair and later Vice-President, and his Digest of Welsh Historical Statistics (1985). He compiled these two invaluable volumes of statistics because of his ‘belief that the quantitative element is a necessary and important part of the historical record; ,,, awareness that it was an aspect that was particularly inaccessible for scholars of Welsh history; and … conviction that some encouragement in the use of quantitative material was necessary’ (Williams, 1985, vol. 1, p. vii).
Some of the references in Williams’s essay may need further explaining in order to help readers understand the essay more fully. His discussion is based on an analysis of the thirteen ‘old’ Welsh counties into which Wales was divided until they went out of existence following local government reorganisation in 1974. He does so of course because the census records for the late nineteenth and early twentieth centuries are compiled in that form (see paragraph 5.2). (In 1995 Wales was divided into 22 unitary authorities following a second reorganisation of local government.) ‘Net loss’ (paragraph 5.5) refers to the size of population lost through migration. It means the excess of migrants out of Wales or a particular county over migrants who settled in Wales or a particular county during the same period. ‘Pastoral’ and ‘arable’ farming in paragraph 5.13 refer to the farming of animals and of the land (growing crops, etc.), respectively. There are a number of speciﬁc references to various aspects of Welsh emigration overseas in the nineteenth century in paragraph 5.14. ‘Patagonia’ refers to the establishing of Y Wladfa (Welsh Colony) in the Chubut Valley in Argentina in 1865. The aim was to create a kind of Welsh state, with Welsh as its ofﬁcial language. In 1890 the USA passed the ‘McKinley tariff’, which imposed high import duties on tin-plate. At the time, the south Wales tin-plate industry was providing the bulk of American imports of the product. The tariff led to substantial emigration of tin-plate workers and their families from the Llanelli–Swansea area, where the Welsh industry was concentrated. The tin-platers found employment in new American tin-plate works in places like Philadelphia and Newcastle in Pennsylvania.
Williams’s essay discusses at some length the work of the economist, economic historian and demographer Professor Brinley Thomas (1906–1994). Thomas is widely regarded as one of the foremost Welsh scholars of the modern era. Born in Pontrhydyfen in the industrial valleys of south Wales, he was Professor and Head of the Economics Department at Cardiff University. He was a world authority on the national and international movements of people and capital, and his works include Migration and Economic Growth (1954). But he is best known in Wales for his thesis that industrialisation in Wales ‘saved’ the Welsh language in the nineteenth century. This controversial thesis has been much challenged but it remains essential reading for the history of Wales in the nineteenth and twentieth centuries.
The poem that Williams quotes from at the end of the essay is ‘The Deserted Village’ by the playwright, poet and novelist Oliver Goldsmith (1730–1774). (Williams had a great passion for poetry.) Published in 1769, the poem laments the disappearance of older ways of life in what the author believed were idyllic rural areas.
Although written about twenty years ago Williams’s essay ‘The move from the land’ remains an exemplary piece of historical writing. It is full of rich insights into the interaction between economy, demography and society in late nineteenth and early twentieth century Wales. Just as impressive is the care and caution with which Williams ﬁrst examines the evidence and then presents his arguments. The essay demonstrates powerfully the author’s extensive knowledge and understanding of his topic and the depth and quality of his skills as one of the ﬁnest historians Wales has produced in recent times.
There are many rewards for attempting to understand the past, but success is not likely to be one of them. This is because the factors which shape the development of economies, societies and nations are inﬁnite, and inﬁnitely varying. Societies may be manipulated (to a degree) by personalities or parties; they may be inﬂuenced by more or less identiﬁable dramatic events; they may be constrained by established institutions and customs; and they may be affected by broad, anonymous movements. Population changes mostly fall within the last category. They steal silently on but may be more continuously signiﬁcant than such more obtrusive considerations as strikes, riots, Lloyd Georges, Liberal parties and Nonconformist chapels.
At the very least, it is perfectly obvious that all aspects of modern history have been crucially inﬂuenced by the unprecedented acceleration in population growth. In the nineteenth century this affected, especially, Europe; in the twentieth century, the world as a whole. One immediate accompanying feature has been the shifts in population as people moved both within nations and between nations. Wales was by no means immune from these broad trends. In 1801 the population of Wales was 587,245; in 1851, 1,163,000; in 1901, 2,013,000. The nineteenth-century pattern, therefore, was for population to double in the ﬁrst half century, and then to (almost) double again in the second half: this roughly reﬂected the pattern for Great Britain as a whole. For Wales, the most dramatic increase came at the end of our period. In the single decade from 1901 to 1911 the population of Wales increased by over 400,000 people. Wales thus, at least fully shared in the general nineteenth-century population explosion. It also experienced the general tendency for population to be redistributed: in the case of Wales this took the form of the balance of the population slipping towards the bottom – in 1801 less than 20 per cent of the population of Wales was in the two counties of Glamorgan and Monmouthshire; by 1911 almos 63 per cent of the people were in these two counties. (Throughout this essay, the old 13 counties of Wales will be used, since this is the form in which all the census material exists.) Only one part of this story will be taken up here. However, that part – the move from the land – was particularly characteristic of the period from 1880 to 1914.
How do we know this? It is not as straightforward as it might seem to identify the Welsh rural exodus. The movement does not show up in vast decreases of population in the agricultural areas. Indeed, in only two counties (Cardigan and Montgomery) was the population in 1911 signiﬁcantly below that of 1851. There are, moreover, no direct ﬁgures on migration in the census returns for this period. Such considerations should make us view statements about migration with care, if not scepticism. None the less, the broad picture can be inferred with reasonable conﬁdence from at least two considerations. In the ﬁrst place we know from the general census ﬁgures that the population of England and Wales, and the population of Wales itself, almost exactly doubled between 1851 and 1911. We also know that the increase must have arisen because more people were born and/or fewer died and/or there was a net movement of people into these countries: but we do not know with any certainty why birth and death rates change or even why people migrate. In other words, the immediate causes of population change are relatively few and simple (births: deaths: migration) but the ultimate causes are numerous and elusive. Fortunately for the present purpose certainty over ﬁnal causes is not necessary. It is enough to notice that the differences in the rates of population change between the different Welsh counties are too great to make it likely that they could be explained simply and wholly by different birth and death rates. Which means that migration must have played a part. Indeed, some of the variation in birth and death rates was itself the result of migration. The movement of people into some areas and out of others would, of course, inﬂuence the age structure of the population of particular areas and this would, over time, inﬂuence the crude birth and death rates (i.e. the rate expressed per 1,000 of total population regardless of the age structure). Even apart from this, however, differences between the rates of population increase in different counties are likely to be mainly due to migration.
This inference is strengthened by the second consideration. It is possible to calculate the amount and rate of natural increase (i.e. the excess of births over deaths) for each county from 1841 onwards. For every county in Wales for every decade up to 1911 this ﬁgure was positive. The population of each county should thus have risen by the amount of the natural increase: in many cases, however, the actual increase was much less (and in some counties for some decades even showed an actual decline). The difference can only be due to people having moved out of those particular counties.
The pattern of net migration which emerges for the Welsh counties from this exercise is quite complex (5Ai and 5Aii ). None the less a few basic points relevant to our problem stand out quite starkly. In particular, starting with the decade ending in 1851, there are ﬁve Welsh counties – Brecon, Cardigan, Montgomery, Pembroke and Radnor – which experienced a net loss by migration in every single decade (5B). Moreover, in four other counties – Anglesey, Carmarthen, Flint and Merioneth – there was a net loss in every decade except one.
These counties then constitute the main losers by migration in the half-century or so before 1914. Before we can conclude that this represented a move from the land, however, we need to know what was happening to the total population of these counties during these decades, and to have some reasonable indicator that the movement represented a shift away from agriculture. The need for caution on the ﬁrst point can be illustrated from the experience of Carmarthen. As already indicated, this county suffered a net loss through migration in every decade except that of 1901–11. None the less, in every single decade there was an increase in the total population. In other words the net migration was only removing part of the natural increase in population. There was movement away but there were always more people left. Whether this constituted a movement from the land is a nice question.
Of course it is possible, even likely, that in Carmarthenshire there was an actual decrease in the agricultural population which was being more than offset by increases in the industrial sectors of the county. Clearly a much more detailed study would be required before such a pattern could be established with certainty: the need for caution is indicated by the broad fact that the number of males in agriculture only fell from 12,900 to 11,300 between 1851 and 1911 (G.B. Longstaff, ‘Rural Depopulation’, Journal of the Royal Statistical Society, LVI, 1893, Tables IV and V illustrate this). The same could be said of other counties in which there were usually net losses by migration but persistent gains in total population. These counties were Caernarfon, Denbigh and Monmouth. Glamorgan was the only county to make a net gain from migration in every decade, and of course it made large gains in total population in each decade. Monmouthshire is particularly interesting since, despite heavy industrialization, it experienced a net loss by migration in four of the seven decades after 1851 although the size of the natural increase (excess of birth over deaths) ensured that the total population rose in each decade. Indeed, the county is in some respects too large a unit to capture the ‘move from the land’. All counties, however seemingly rural, have urban centres and pockets of industry or mining. Some of the complications of this have just been indicated (an attempt to deﬁne the rural population more precisely is given in A.L. Bowley, ‘Rural Population in England and Wales’, Journal of the Royal Statistical Society, LXXVII, 1914). But in other important respects, the crudity of the county as a unit is an advantage. The essence of the ‘move from the land’ as an issue in Welsh historical experience is not a concern about shifts within a county but about a fundamental change in the balance of the population between different regions of the country as a whole.
At all events, the real hall-mark of the counties where out-migration was the most characteristic demographic feature was: a persistent net loss of population by migration accompanied by frequent decades in which there was also an absolute decline in the total population. The counties so affected were Brecon, Cardigan, Merioneth (for our particular period from 1800 onwards), Montgomery and Radnor. These were the counties which experienced an unequivocal outwards movement. Was it also a movement from the land?
The most useful indicator of that would be to see if these counties were primarily agricultural. And that can be best shown by looking at the proportion of the total working population engaged in agriculture (5C). On this basis it turns out that the counties where out-migration was most marked were those which were most heavily agricultural. Even in 1911, when for Wales as a whole less than 12 per cent of occupied Wales worked on the land, Radnor, Montgomery and Cardigan each had around one-half of their male workers in agriculture (Merioneth had one-third, and Brecon nearly one-quarter).
What has been stressed so far is that population movements are complicated and often confusing, especially in the absence of any direct census information on migration. None the less it has been more or less conclusively inferred from the evidence that there was a substantial redistribution of the population within Wales and that one signiﬁcant feature of these shifts was a movement of people away from the countryside. It is still not possible to identify with any certainty just who moved, but there are strong indications that one element was an outﬂow from agriculture itself. In the ﬁrst place, the total number engaged in agriculture in Wales, despite a doubling of the total population, actually declined between 1851 and 1911 (most of the decline being in the number of farm workers and not in the number of farmers). Even more suggestive is the fact that the rate of decline was generally most steep in the more agricultural counties (5D). In ﬁve counties (Brecon, Cardigan, Merioneth, Monmouth and Montgomery), the number of males occupied in agriculture in 1911 was less than three-ﬁfths of the number for 1851. Four of these counties were heavily agricultural, the oddity again being Monmouthshire. An important group contributing to the out- migration was clearly made up of those who had been, or would have been, occupied in agriculture.
The question of who moved raises another, and perhaps more interesting, issue. There is a strong general presumption that the rural counties lost the most able, energetic and industrious section of their population. But did they? The question is fundamentally intractable. The assumption that the best would move is plausibly based on the proposition that only the most able would both see, and seek out, the opportunities opening out in other regions (5E). Still, the draining-away of those of the highest quality ought not simply to be presumed. It is also plausible, perhaps equally plausible, to argue that in a declining, highly competitive agricultural environment only the best would come to own farms and make them pay: the rest would have been driven out. It is often those who have been least successful in a given environment who are squeezed out and this should at least call into question the presumption that the agricultural areas are necessarily losing their most able people. What is true, of course, is that the migrants tended to be concentrated in the younger age groups. Indeed, this is in itself one reason for the general assumption that the agricultural areas lost their most able people: in the context of the nineteenth century it was quite reasonable to confuse strongest with best.
Another aspect of this which on the face of it seems a likely cause of movement was that technical change in agriculture was reducing the need for labour. This is given additional plausibility by the fact that the movement out of agriculture was mostly a reduction in the number of agricultural labourers and not in the number of farmers. None the less, the use of machinery seems to have played little part in this process (5J). Most of the reduction took place because one of the effects of foreign competition was to drive Welsh agriculture towards heavier concentration on pastoral as against arable farming: between 1867 and 1914 the total acreage under permanent pasture in Wales increased from 1,500,000 to 2,300,000, whilst that under arable cultivation fell from 1,100,000 to 700,000 acres.1 One consequence of this was a reduction in the demand for agricultural labourers because in general pastoral farming is less labour-intensive.
Brinley Thomas also provides a powerful and cogent economic explanation for the pattern which emerged. This need not be repeated here since the immediate concern is simply with the question of the destinations of those who moved from Welsh rural areas. From the 1800s they mostly stayed in Wales. There was always movement out – to England and to countries like the United States: but these had ceased to be the typical destinations. The movement was toward the southern coalﬁelds which were absorbing labour at a very high rate, especially in the 1900s, and also, more modestly, towards the growing holiday towns of north Wales.
The substantial population shifts which have been indicated naturally had far-reaching social, economic and political repercussions. They reﬂected the considerable change which took place in the economic structure of Wales. In economic terms this contributed to a general increase in income because it represented a move of resources away from low-productivity agriculture and towards higher-productivity industry.
In 1851, 35 per cent of all occupied males were engaged in agriculture: in 1911 only 12 per cent were so engaged. The absolute decline in the number occupied in agriculture was not all that great (from 135 to 96 thousands), but it meant that the industry had not been able to absorb the natural increase in the population which thus had to look for employment elsewhere. In the case of those counties which were primarily agricultural this usually meant leaving the county altogether. The transition of Wales from an agricultural to an industrial nation was effectively accomplished during these years. One of its consequences was that the population became, by 1911, much more concentrated around the south Wales coalﬁeld, and around the slate quarries and smaller coalﬁeld in north Wales.
One of the features most generally commented upon – both at the time and since – was that it was a process which involved rural depopulation. And so it did – of a sort. But it is important to be clear about the nature of this depopulation. Brieﬂy it can be said to have had three broad characteristics: it was highly concentrated; it was limited; and it was most conspicuous as a relative phenomenon. Five counties5 experienced some fall in the absolute level of their population in the half century or so before 1911 and these were, as already seen, mainly the counties in which agriculture was most important. Indeed, between 1881 and 1911 the actual decline was concentrated in only three counties – Cardigan, Merioneth and Montgomery – where the combined population fell from 188,000 to 157,500. The fall was by no means catastrophic. The total population of the ﬁve counties in 1881 had been 269,300 and in 1911 it was 240,500. But although the loss of rather less than 30,000 people over three decades cannot be seen as dramatic depopulation, it has to be given perspective in at least three broad ways. First, the late nineteenth century was simply the beginning of a long-run process. The movement from the land set in train an extended, creeping depopulation from, especially, rural mid-Wales. By 1971 the total population of the ﬁve counties had fallen to 205 thousands. The second necessary perspective is to view the population decline in these counties against the contemporaneous trends in other parts of Wales. In ﬁve of the most industrializing counties6 the population increased between 1881 and 1911 by well over one million. The two counties of Monmouthshire and Glamorgan on their own accounted for nearly 800,000 of this increase. And ﬁnally, it needs to be noted that the rural population exodus affected a much higher proportion of the counties of Wales than of England and to a much more marked extent. Thus between 1881 and 1891 most (eight) Welsh counties suffered a population decline compared to only twelve of the English counties, and in ﬁve of the Welsh counties (mostly in mid-Wales) the percentage decline was greater than even the worst-affected English county (5P).
The effect of the move from the land on the health and viability of the Welsh language has been much more controversial. For long the conventional wisdom was that it was part of a process which necessarily undermined the language. Welsh-speakers were being drained away from rural districts where the use of the language was most general and where it was most protected from outside inﬂuences. It was often an implicit assumption of this view that these Welsh- speakers emigrated to America or Patagonia or wherever. But that aspect was not ultimately essential since an even stronger assumption was that industrialization necessarily undermined the language and its mutually-supportive culture. This entrenched position, however, became untenable after it had been subjected to the decisive bombardment embodied in Brinley Thomas’s seminal article on ‘Wales and the Atlantic Economy’.7 It was demonstrated, as already seen, that most of the Welsh migrants were able to stay in Wales and, moreover, Wales kept most of its natural increase in population. And all this was only possible because of the industrialization of Wales.8 Thus the total number of Welsh-speakers in Wales was almost certainly higher than it had ever been before. And, it is argued, that this – despite the fact that they constituted a smaller proportion of the total population – ‘gave the Welsh language a new lease of life’ and ‘a second chance’ to establish itself (5Q).
The snag was in the ‘several times over’, for that meant that there was necessarily a huge inﬂux of non-Welsh. This was on a scale (at least 100,000) and in a time-span (a single decade) that made absorption unlikely or impossible. In addition the huge growth of the mining labour force in this decade makes it clear that this particular tidal wave was not conﬁned to the already strongly anglicized coastal ports of Newport and Cardiff and Barry.
Much of this has been recognized and, indeed, presented by Brinley Thomas in his latest work.10 But it is still robustly maintained it was only the industrial revolution and economic growth which prevented the Welsh language (and people) going the same way as the Irish; that the million or so Welsh speakers at the end of the nineteenth century gave Welsh a second chance; and that, contrary to myth, the life and rigour of the culture was centred in industrial Wales and not in the so- called rural heartland. It is an issue that will continue to be debated but is not likely to be resolved because, like many aspects of the move from the land, it manifests itself in ways both immeasurable and imponderable.
Much the same might be said of the effects of emigration on social customs or the elusive, but signiﬁcant, concept of the idea of the community (5Ri, 5Rii). In part, therefore, judgements will necessarily be subjective and depend on how individuals respond to Goldsmith’s assertion that
‘A bold peasantry, its country’s pride,
When once destroyed can never be supplied.’
Debates about the effects and seriousness of the move from the land, in other words, are conditioned by the extent to which one subscribes to some concept of a rural arcadia. Some people see rural life as having inherently superior values; others see it as having simply more space.