As Christopher Smout shows in Chapter 1, for its size, Scotland has a very diverse natural environment. Historically, there have also been major regional differences in landholding practices, industrial development, and popular and religious culture. All these produced a highly differentiated spatial demography; and this means that we always need to go below the national level if we are fully to understand Scottish population change and its implications for the people of Scotland.
In practice, however, we have extremely little hard evidence about regional or even national populations before 1801. Robust information on births, marriages, and deaths is only available from 1855. Migration remains a problem even in 2010.
Scotland’s first official census, in 1801, showed a national population of somewhat over 1.6 million, with figures for every civil parish. In spite of some problems with data collection, the results for most places are probably accurate for the civilian population to within a few per cent.
By contrast, though widely used by modern historians, Alexander Webster’s estimates for 1755 require considerable caution, especially at the local level.1 Webster did challenge some figures from parish ministers, but many were only estimates. Frequently, ministers only provided rough guesses, especially in the towns, whereas others gave numbers they considered eligible for ‘examination’ on the Catechism, plus the age at which they treated children as ‘examinable’. Webster implies that he estimated children under the examinable age using a standard formula, derived from Halley’s Breslau life-table. But Scotland was not Breslau, and variations in mortality and migration meant that age structures between parishes varied widely, so Webster’s method could only produce very approximate results at the local level even if he applied it systematically. But in fact he did not, instead frequently adjusting many of the ministers’ figures, presumably to make the results look more plausible by removing obvious roundings to zero. One result is that well over a quarter of his parish totals have final two digits in the range ‘89’ to ‘99’, two and a half times the randomly expected number. But the surviving returns from ministers suggest no such pattern in what was sent in.2
Use of Webster’s parish figures thus requires a strong degree of scepticism, especially in towns and the Highlands, though, at larger levels of aggregation, some errors will cancel out. The most likely conclusion is that Scotland’s 1755 population fell somewhere between 1.15 and 1.35 million, possibly towards the higher figure—but, even at 1.2 million, Scotland’s growth to 1801 would have been significantly slower than England’s.
Most population estimates for earlier dates are largely guesswork.3 Robert Tyson suggests a total of 1,234,575 for 1691.4 But this derives from incomplete counts of hearths from tax returns, adjusted by regionally varied estimates of the number of hearths per household, plus data on mean household size by county from a century later. Given these levels of uncertainty, we should not even try to guess more than very rough national population figures before 1801, but focus on what we can infer, from contemporary commentary or data on related matters, about the likely impact of changing opportunities and pressures on people’s experienced demographic lives.
On this basis, what might we suggest? Older work mostly assumed that variation in mortality was the key source of demographic change, in a high-pressure world of elevated birth and death rates. So, for example, it was suggested that inoculation against smallpox provided significant stimulus to population growth in the later eighteenth century. Unfortunately, the timing and extent of inoculation makes this unlikely except in the Highlands and the south-west (where there is also supporting contemporary comment).
Above all, however, great stress was laid on fluctuations in death rates, based on contemporary comment and, for the later seventeenth and eighteenth centuries, on counts from parish registers.5 These suggested that recovery from devastating disease and famine crises of the early fourteenth century was delayed by recurrent outbreaks of bubonic plague before 1500, implying a national population in 1500 of 500,000 to 700,000. Plague and crop failure recurred frequently throughout the sixteenth century, but the number of very high food-price years, plus increasing action against vagabonds, hint that the population rose significantly by 1600, perhaps even to one million, but that it probably grew little if at all thereafter.
Nor, in Scotland, did mass mortality linked to harvest failure end after 1600, as it largely did in England, except for the north-west in the 1620s. Seventeen Scottish dearth years in the seventeenth century included major mortality crises in 1623–4, 1650–1, 1673–6, 1681–3, and, above all, in the 1690s, when probably at least a tenth of Scotland’s population died—but perhaps as much as a quarter in inland areas of the north-east and much of the western and northern Highlands. Later still, deaths from hunger-related disease were reported in the 1740s and the 1790s, especially in the Northern Isles and the northernmost mainland.6 Elsewhere, improved market integration, better transport, and proactive steps by local landlords or Poor Law authorities reduced hunger migration and its consequential deaths from famine-related disease (though we still need further research on these developments, from the 1690s onwards).7
Locally, before 1700, short-term crises could be severe, a key factor in the constant insecurity of everyday life at this period. In really bad crises, a third of a parish’s population could die within a few months. However, most demographers now believe that the medium- and long-term impact of crises on population trends anywhere in north-western Europe was less significant. The only exceptions were crises of exceptional severity (as in the fifteenth century, or in Aberdeenshire in the 1690s, where some parishes took several decades to recover8), or if crises followed each other closely clustered in time (as happened in Scotland in the civil-war years).
The main reason for rapid recovery was that most areas in north-western Europe had many young adults quite literally ‘waiting’ for an opportunity to marry, through land, or another regular source of income, becoming available. When the opportunity arose, because most new households were headed by younger couples, fertility was also boosted. In Scotland, delayed marriage seems to have been normal in lowland farming areas before 1800, with an average age at first marriage for women in their later twenties, and perhaps one in ten not marrying in their fertile years. Scottish Lowlands thus had capacity for rapid recovery after a mortality surge; the same was true in the north-west and the islands, though marriage ages were often rather lower here.9 Lowland populations in particular were also highly mobile, so mortality-induced gaps in one community were readily filled by in-migrants from elsewhere.10
Most of the time, these processes restricted population growth. For example, Tyson11 shows that, between 1755 and 1841, many landlords in the north-east and central Highlands not only limited additional household formation by prohibiting subdivision of holdings, but consolidated holdings when the opportunity came up, shifting from mixed farming to stock rearing. In one parish with reasonably robust data (Rothiemay in Banffshire), mean marriage ages rose by over two years for women and four years for men from 1750 to 1800 and 1801 to 1851. In the Western Isles, a similar policy restricted population growth on Coll to less than 10 per cent between 1755 and 1831.
But elsewhere on the west coast and in the islands, restrictions were widely relaxed, especially in the later eighteenth century, as landlords encouraged or even forced holding subdivision to encourage new marriages and in-migration on the coast. The extreme was Tiree, where these policies were reflected, as early as 1779, in a mean female marriage age around 22 and over a quarter of women marrying before they were 20. In 1799, as the kelp boom took off, larger farms were also subdivided and in some areas in-migration was actively encouraged. Population roughly trebled between 1755 and 1831, and subdivision only ceased in the 1840s. Similar processes were widespread around the Argyll coast and the islands at least through to 1815.12
However, the consequences were severe when the market for kelp collapsed after 1815, and fishing became increasingly unreliable. In many parishes, populations had only been marginally viable at much lower levels even before the kelp-related boom. Only the potato had made the new numbers sustainable under any contemporary agrarian regime—and, when the potato crop failed dramatically in the 1840s, massive population decline in parishes like Tiree was inevitable, whatever landlords might have sought to do.13 But, beyond this, ever since the mid-eighteenth century in particular, some landlords in a growing number of areas of Scotland had been actively reducing even populations with more robust means of subsistence in favour of larger- scale arable or livestock farming. The ‘Highland clearances’ of the first decades of the nineteenth century were merely one part of a longer and much more widespread trend.14
Tyson also notes a very contrasting picture, illustrated by a quote in the New Statistical Account from the minister of the fishing parish of Rathven in Banffshire; this also has parallels right across western Europe in this period. ‘At eighteen years of age they become men, and, whenever they acquire the share of a boat, they marry, as it is a maxim among them “that no man can be a fisher, and want a wife.”’15 The result, it was claimed, was almost universal male marriage, and women typically marrying between 18 and 22. Elsewhere in western Europe, expansion of employment in low capital-base craft and manufacturing activity and in mining created similar opportunities for couples to establish their own homes at earlier ages than in the past, thus stimulating local population growth. New opportunities also encouraged in-migration from rural areas where marriage chances were more limited. The differential regional population trends of the years 1755 to 1851, to be discussed further below, are entirely consistent with this kind of behaviour also occurring in Scotland.
It seems therefore that, as in England and many other countries, opportunities for marriage were a major control over Scottish demographic change in this period. However, in the decades around 1800 there is no sign in Scotland of the significant fall and subsequent rise in mean marriage age that was widespread in England. Given the similar economic and social changes in the two countries by this time, this is surprising, and further research is needed if we are to understand it. However, two areas that were important in the nineteenth century may also have been at work earlier.
First, in early nineteenth-century lowland Scotland, a much higher proportion of agricultural labour than in England was accommodated by farmers, and contracted for at least six months at a time—and there were fewer sites for squatters to settle; the result was greater control over how many single or married people could reside in each parish. Second, the Poor Law in England, especially from the later eighteenth century until 1834, supported casual labourers and their families in their own homes when work was not available, and even subsidized families on the basis of numbers of children. In Scotland, by the mid-nineteenth century, no significant Poor Law support was provided to indigent workers and their families; this was arguably an important risk factor restraining Scottish nuptiality at this time. More work is needed to see whether this also applied in really bad times in Scotland in the eighteenth century, but Poor Law support in many parishes was anyway much less effective at that time.16
Finally, what about migration?17 Even in the medieval period, Scots were renowned in Europe as mobile people. From the sixteenth to the mid-nineteenth centuries at least, much internal migration was of younger men and women moving through a series of short hirings between rural communities as farm workers, or spending time in trading or manufacturing activities, especially in the towns. These needed migrants to grow—and in this period, especially in the larger ones, to counteract their high death rates (though up to the mid-eighteenth century the overall demographic impact of urban centres was limited in Scotland compared with many other parts of Europe, with less than 2 per cent of the population living in towns of more than 10,000 people in 1500 and only about 9 per cent in 1750—thereafter was a different story).18 Until the early eighteenth century, there were also many famine migrants and vagrants in bad times, large numbers of whom died (mostly from disease) as a result. Later, as industrialization and agricultural improvement accelerated, with an accompanying rise or decline of local opportunities in farming, fishing, or manufacturing, growing numbers of families moved in response, which was reflected, for example, in the widespread falls in population in many north-east parishes in the later eighteenth century.
Externally, especially through the sixteenth and seventeenth centuries, many Scots traders went to parts of Continental Europe (more than went to England). Scots also formed major components of continental mercenary armies (over 50,000 on one estimate in the 1620s and 1630s alone)—and many of these never returned. And there were major outflows to Ulster at various points in the seventeenth century (and notably in the 1690s). Overall, probably about 200,000 Scots left the country in the seventeenth century, enough, at an average of two per hundred population per year, to provide an important supplementary safety valve against excessive population pressure.19
Overall in the eighteenth century, net emigration was lower as a proportion of national population, not least because, by its end, there was growing net inflow from Ireland, especially into the south-west and the west Central Belt.20 Emigration probably reduced population growth between 1750 and 1800 by less than 5 per cent, but regionally there was real impact in parts of the north-west, with perhaps 20,000 emigrants to North America between 1765 and 1815 coming from the Highlands, where the average population was well under 200,000.21 Even more than earlier, in this period many men also moved or were moved out of Scotland as soldiers (up to a fifth of adult males in some areas of the north-west by 1805), as pressed sailors, and, especially in the Northern Isles, in fishing and whaling—and, in all these categories, many died abroad.
After 1800, information on key aspects of Scotland’s population becomes more robust. National populations from the decennial censuses, and changes compared with other parts of Britain and Ireland, are shown in Table 2.1. Census data for 2011 are not available at the time of writing and, given the many uncertainties about the pattern and causes of Scottish demographic change after 2001, discussion of the years 2001–10 is here confined to a short final section of this chapter.
Table 2.1 Scotland’s population, and decadal rates of change for the components of Great Britain and Ireland, 1801–2001
Compared with England, the Scottish population grew more slowly in every decade between 1801 and 2001. Scottish growth was a little over threefold, England’s was nearly sixfold, and Welsh nearly fivefold; only in the 1870s and 1890s did Scottish growth approach England’s. From 1921 to 2001, Scotland’s population fell in four decades; England’s lowest growth was a 1.7 per cent rise in the 1970s. Wales matched Scottish rates in the later nineteenth century, grew rapidly in the 1900s, fell in the interwar period, but rose much faster after 1951. Ireland’s population boomed from the mid-eighteenth century until 1841, but the failure of potato crops initiated nearly a century of decline, followed by modest growth, though in the 1970s and 1990s Ireland grew much faster than any country of Great Britain. Even compared with English regions with a large concentration of staple industries, Scotland’s growth was in general slower, except marginally for the English North and North-east in the first decades after 1945.
In a wider European context, Scotland’s sluggish performance is even more striking. Up to 1911, Scotland’s growth, though markedly slower than England’s, was much faster than France, and not very different from Belgium, Switzerland, or Sweden, but significantly slower than Denmark, Norway, the Netherlands, or Germany. But, after 1921, Scotland’s relative demographic depression dramatically stands out. No significant Western European country, regardless of size, saw decades of actual population decline in the inter-war period. Scottish population rose by only 2 per cent between 1921 and 1951, while that of Belgium, Sweden, Denmark, and Norway increased by between 17 and 41 per cent. The last half of the century was relatively even worse. Scotland’s population in 2001 was about 32,000 less than in 1951; everywhere else in north-western Europe, populations grew, in some cases rapidly. The fastest growth was in the Netherlands (over 50 per cent), but Sweden, Norway, and Denmark all exceeded 20 per cent.
Just why Scotland’s population was so relatively sluggish for so long cries out for further research, but part of the answer is that there were multiple Scotlands. Published census material at parish level is available from 1801. There are awkward boundary changes in the 1890s, and problems as urban areas extended their boundaries and from local-authority reorganizations in 1975 and 1996. Nevertheless, we can reconstruct six reasonably consistent regional groups; the results are shown in Figure 2.1.
FIGURE 2.1 Populations of the regional groupings of Scotland, 1755 to 2001 (in thousands).
First, note the huge changes in the relative shares of the different regions of Scotland. For 1755, Webster’s figures suggest that the Highlands and Islands had over a fifth of Scotland’s population, at a time when Strathclyde (less Argyll, which is here treated as Highland) held only 15 per cent, and the Lothians 10 per cent. Thereafter, for two centuries, the developing industrial and commercial activities of the west of Scotland meant that Strathclyde’s share rose almost continuously, peaking in 1961 at nearly 49 per cent, before falling back below 43 per cent by 2001 (the fall in Glasgow was even larger). In 2001, by contrast, the Highlands and Islands held just 6 per cent of the population (and the East and West Borders together a mere 5 per cent—compared with nearly 12 per cent in Webster’s figures).
But if we look at absolute numbers, a rather different picture emerges. The 2001 Highlands and Islands population was around 315,000, up from 298,000 in 1951 and above the 306,000 of 1801 (even if Inverness is excluded, the 1801 figure is still around 297,000 and that for 2001 about 271,000). The Borders, with a population almost identical to that of 1841, had over 100,000 more people in 2001 than in Webster’s estimates for 1755. Strathclyde’s population in 2001 was eleven times that of 1755 and more than six times that of 1801, but was 15 per cent down on its 1961 peak. No other major manufacturing region in Europe experienced anything approaching this scale of loss, and Strathclyde is a principal cause of Scotland’s comparative demographic weakness in the twentieth century.
But even these figures conceal subtler changes in Scotland’s population. In 2001 most Scots were spatially highly concentrated, with 63 per cent living in parishes with a population density above five hundred persons per square kilometre (km2); these parishes occupied just 3.3 per cent of the land area, mostly in the Central Belt or around Aberdeen, Perth, or Inverness.22 At the other extreme, 46 per cent of the land area consisted of parishes with fewer than five people per km2, holding just 1.7 per cent of the population. And in these sparsely populated parishes, excepting some coastal crofting areas, most residents were concentrated into villages or other grouped settlements, thus further boosting the image of the remaining ‘deserted lands’.
But, as Map 2.1 suggests, most of these parishes had always been sparsely settled, the only major exceptions being some island parishes and parts of Argyll and Caithness—but even these areas were only settled at a fraction of the density of most of rural Ireland in the years before the famine (for example, even Tiree, which was among the most crowded, had a density in 1841 that was only a third of the average for the County of Cork). Except for a few fishing centres, there was hardly a mainland parish on the north-west coast of Scotland between Fort William and Thurso, and also deep into western Perth and Aberdeenshire, where population density in 1801 exceeded ten per km2; mostly, the average density was much less. Durness in Sutherland where just 1,208 people were settled on 570 km2, had an average density of 2.1 people (less than half of a household) per km2. These low densities arose because, even within highland parishes, most areas were largely empty excepting only parts of the most fertile glens. At the other extreme, the principal clusters of population by 1801 were already located in the Central Belt, up the east coast, around the Moray Firth, and along the coast of Dumfries and Galloway. Two centuries of spatial change, therefore, principally made sparsely populated parishes sparser, and densely populated areas denser—and the same process was also at work within most parishes, as noted above.
MAP 2.1 Population density (persons per km2), Scottish parishes, 1801.
But the process of population decline in the now sparsely settled areas was not homogeneous. Map 2.2 shows the widely varying dates, even between neighbouring rural areas, when parish populations peaked.
Some explanations for differences are well studied: the contrast between Argyll and much of Perthshire, where landowners were consolidating holdings (initially mainly for cattle) in the later eighteenth century, compared with eastern Aberdeenshire, where new settlement was actively encouraged for small cattle-breeding farms even in the 1870s; or the mixture of active clearance and flight from deprivation in the Uists, Skye, and Sutherland before 1851, compared with Harris and Lewis, where prosperity from fishing partly backed by landlord investment delayed most decline until the later nineteenth century; or the contrasting fates of smaller and larger east-coast fishing ports, as processing concentrated in those with the largest fleets and best communications.23 But under-researched puzzles remain: for example, why did almost every parish in Wigtown peak in 1851 (it was more than just an Irish effect); why was decline later in north Berwickshire than in eastern East Lothian (was it because of the water-powered industry in the Whiteadder valley?); and what population effects did turnpikes or railways have?
Map 2.2 also shows that many of the industrial, mining, or commercial parishes that had expanded so rapidly between the later eighteenth century and the First World War, then subsequently peaked long before 1971.24 This included a swathe of mining districts across south Ayrshire, and, as their mining, shipbuilding, iron and steel and engineering industries went into decline, especially after 1921, most of the parishes north-west and south-east of Glasgow, including the city itself. Less dramatic falls can also be observed after 1951 in the city cores (though not the suburbs) of Edinburgh, Aberdeen, and Dundee.
MAP 2.2 Peak population decades for Scottish Civil Parishes, 1801–2000.
How do we understand all this? Before we can link civil registration to the censuses after 1861, it is difficult to be precise about the causes of Scotland’s sluggish demographic change. Very high urban mortality from the late 1820s, as rates of town and city growth unparalleled in Europe overwhelmed housing and sanitary infrastructure,25 was a factor, but more detailed local studies outside the cities are needed to establish its national importance. In spite of the major Irish inflow, net emigration was high in the 1830s and 1840s, as manufacturing experienced the first major trade-cycle depressions, and also in the 1850s (when it spread right across the country). But why Scotland fared much worse than England is still not clear.
After 1861, we can explore for each decade the precise net interplay between numbers of births, numbers of deaths, and net migration. The results are shown in Figure 2.2.
FIGURE 2.2 Components of Scottish population change, 1861–2001.
These data have been explored extensively elsewhere.26 For present purposes note: the steady decline in natural increase of population from the 1870s to the 1930s, as the birth rate fell faster than the death rate; the stabilizing death rate from the 1950s, as improved survival was offset by an ageing population; and the boom and then collapse of the birth rate, which eventually pushed Scotland into negative natural increase in the last decade of the twentieth century.
In many ways, these trends roughly parallel those of England and Wales, and, indeed, many other parts of Western Europe. Marital fertility began to decline significantly in many parts of most countries from the 1860s at the latest, for reasons that we still do not fully understand. Scotland’s pattern is fairly typical: a shift to small families, with their huge implications for living standards and the lives of women, was widely occurring by 1911 among professional middle-class and textile workers (including a growing number of very small families of zero and one), while many large families persisted until the 1950s in mining, heavy industry, and crofting areas—though early fertility decline in farming areas such as East Lothian (where many women worked in the fields) appears in comparative terms unusual.27
But two features are internationally special about Scotland’s demography over the 150 years to 2001: first, the rate of net emigration (including movement to other parts of the UK); and, second, the substantial variability of birth, marriage, death, and migration rates in different parts of the country.
In the 1880s and the first decade of the twentieth century, net out-migration exceeded 5 per cent of the population and remained high right up to the First World War; between 1904 and 1913, more than 600,000 people left the country (annually more than 1 per cent of population in six of these years and more men than were killed in the First World War). Overall, between 1830 and 1914 about two million people left Scotland for overseas destinations and probably roughly similar numbers went to the rest of the UK, making Scottish emigration per head second in Europe after Ireland. In the 1920s net out-migration exceeded 8 per cent of the population and was the principal factor producing real population decline; some 400,000 people (net) left Scotland in these years, mostly, after the introduction of severe restrictions on migration to the United States in 1923, to the rest of the UK. Net outflow was high again in the 1950s (around 5.5 per cent) and the 1960s (around 6.3 per cent), in total some 600,000 people, about half going overseas. Continued net out-migration at these levels would have produced massive population decline after 1970, but, as natural increase fell, net out-migration declined, to just 2.9 per cent in the 1970s, 2.4 per cent in the 1980s, and 0.3 per cent in the 1990s.28
The situation in England was dramatically different. In spite of high overseas emigration in terms of numbers, England had relatively low rates of net loss even in the first thirty years of the twentieth century (the highest was 1.7 per cent in 1911–21), and migration was mostly modestly positive thereafter (between 0.8 and a little over 0.9 per cent per decade between 1921 and 1951, and again between 1981 and 2001).
But Scotland also attracted large numbers of in-migrants, actually having a higher proportion of non-natives than England back to at least 1851. The big difference in Scotland’s situation was that most Scottish immigrants after the Second World War came from elsewhere in the UK, while a growing majority of those south of the border came from overseas. Thus, in 2001, 9.1 per cent of the population of Scotland was born in the rest of the UK (up from just 5.4 per cent in 1951),29 whereas in England and Wales this number was just 2 per cent. However, most of this 2 per cent consisted of Scots, the equivalent of 16.2 per cent of Scotland’s population in 2001 (up from 11.4 per cent in 1951 and 7.1 per cent in 1901). In 2001 England and Wales had 8.9 per cent of their people born outside the UK; for Scotland it was just 3.8 per cent (though this was significantly above the 2.5 per cent of 1951).
The other remarkable feature of Scotland’s net out-migration has been its spatial pervasiveness. When this can first be measured, in the 1860s, over 90 per cent of Scottish parishes, holding over 60 per cent of the 1861 population, had net out-migration. This included all but nine parishes in the Highlands and Islands and Perthshire, all but six in the north-east, and most of the farming parishes in the southeast and south-west, where rates of loss were often far higher than from the Highlands and Islands.30 But, even this early, it also included many urban and industrial parishes, among them nine with populations of more than 20,000, the largest of which was Paisley. More remarkably in a UK or European context, after 1870 at least one of Glasgow, Edinburgh, Dundee, and Aberdeen (by the 1900s all the four) experienced net outflow in every decade.
The other defining feature of Scottish demographic history right up to the 1970s was a remarkably repressed nuptiality, especially for women.31 Not only was there a greater reluctance to marry in Scotland than in similar kinds of areas in England, but, in the Highlands and Islands, Scottish nuptiality was among the lowest in the whole of north-western Europe. Possible explanations for this, but with even clearer applicability for the century after 1850, have been discussed above, to which we may add markedly lower real wages than in England, as a discouragement to all but those most confident about their futures from entering into marriage.
Nevertheless, while nuptiality was repressed in the century before the Second World War compared with south of the border, marital fertility was somewhat higher, though also regionally variable. Figure 2.3 illustrates this for different groups of Scottish counties for 1911 and 1961, and compares them with English counties as a whole. In the absence of detailed information on ages at childbirth, this graph uses the Princeton measure of marital fertility (Ig) and relates it to the Princeton measure of nuptiality (Im). Ig is an age-standardized proportion of ‘potential’ (and roughly ‘maximal’) fertility that a group of women actually produced over a ten-year period around the census date. Im is most easily interpreted by noting that 1-Im gives the proportion of a woman’s potential fertility that is not available for conception within marriage, because of delayed marriage or non-marriage.32
FIGURE 2.3 Age-standardized nuptiality and marital fertility, Scottish and English counties, 1911 and 1961.
The pattern for 1911 shows the highly characteristic features of Scottish nuptiality and fertility before the First World War. Firstly, compared with England, Scottish nuptiality tended to be lower, but, for reasons that we do not yet understand, marital fertility significantly higher; as a result, overall levels of fertility were similar. Second, Scottish fertility and nuptiality were spatially more differentiated than in English: notably, the Highlands and Islands had extremely low nuptiality but high fertility within marriage, the industrialized and mining counties of the west Central Belt tended to be high on both (with West Lothian as the UK peak on both), and Roxburgh, Selkirk, and Peebles, with their strong concentration of female employment in textiles, were low on both.33
The 1961 pattern was different, even though, overall, age-standardized fertility in Scotland and England had only fallen by 10 and 9 per cent respectively (it had been markedly lower in the 1930s). But, if we break this down into its components, nuptiality was much higher in both countries and marital fertility much lower—yet, for reasons we still do not fully understand, Scotland still tended to rather lower nuptiality and higher marital fertility than England. Within Scotland, a similar regional patterning to that of 1911 was still visible, but with a markedly reduced range of variation.
One caution, however, has recently emerged: in the north-west, high marital fertility may partly reflect a custom whereby women living elsewhere returned to their mothers’ homes to bear their first child.34 The same behaviour also partly explains high non-marital fertility in some parishes, notably in the south-west and parts of Banffshire and Aberdeenshire, where over a fifth of babies were born to non-married mothers in the later nineteenth century.35 This raises an important issue for further research: it could well be, for example, that highland-born migrant women were as likely as those born elsewhere to marry somewhere, but this is concealed by our only having evidence on the limited numbers who remained at home. The same may also apply to women born in Wigtownshire, for example.
In the later twentieth century, significant changes occurred in both nuptiality and fertility, both within and outside marriage. First, the mean age at first marriage for men, which was about 27 in the late 1850s and peaked at 28.2 in the 1930s, fell after the Second World War to a low of 24.2 in the early 1970s, then rose to 30.7 by 2001. A parallel pattern for women produced a peak of 25.4 for 1911–20, a post-war low of 21.8 in the mid-1970s, and a rise almost to 30 by 2001.
Second, for both men and women, the proportions who married rose markedly in the post-1945 period and fell dramatically in the late twentieth century. In 1971, 58 per cent of women aged 20–24 were married and 91 per cent of those aged 30–34. This compares with 22 per cent and 67 per cent in 1911, and 8 per cent and 61 per cent for 2001.
However, much of the post-1970s marriage decline was substituted by a rise in cohabitation. Cohabitation was very uncommon even among the childless as late as the mid-1970s but, in 2001, despite reduced levels of marriage, a quarter of women aged 20–24 and three-quarters of the 30–34 age group had either at some time been married or were cohabiting, figures well above the proportion who had ever been married in Scotland for any year before 1939 (when cohabitation was extremely uncommon). In the long historical picture of partnership, it is the post-1945 period and not the late twentieth century that is the anomaly.
The overall trend in fertility was markedly downward in the later twentieth century. However, this significantly reflected not a flight from childbearing altogether but the near disappearance of very large families, plus a general shift to later childbearing, which itself resulted in more very small families. As late as 1981–5 women’s median age at childbearing was 25; by 2001–5 it was 29.3. In the same period, births to non-married mothers rose from 15 per cent to 45 per cent of all births (Dundee’s figures were 25 per cent and 60 per cent respectively). Most of this rise was directly linked to the growth in cohabitation. The first detailed data, for 2001–5, show almost two-thirds of all births to non-married mothers as registered by couples living at the same address and just 14 per cent by a mother alone. More generally, however, the whole question of changes and regional differentiation in fertility is still poorly understood even for twenty-first-century Scotland, and there remains great scope for more oral history as well as statistically focused research for earlier periods.
Detailed examination of mortality changes in the nineteenth and twentieth centuries is beyond the scope of this chapter; indeed, most of the mass of mortality statistics produced by the General Register Office has not yet been addressed by historians. But a few key points are clear. In the early twenty-first century, Scotland had the lowest life expectancy at birth in Western Europe. Within the United Kingdom’s 432 local authorities (LAs) in 2005–7, only three Scottish authorities were in the best 200 for women’s life expectation either at birth (e0) or at age 65 (e65), and six Scottish LAs in Greater Glasgow were among the worst ten UK LAs for female e0.36 In 1999–2001, Glasgow City was four years below the UK average figure of 80.2, and 2.3 below the Scottish average.
Over the long term, however, there are some brighter signs. Over the previous 150 years, e0 rose dramatically from 40.3 years for men and 43.9 for women in 1861–70 to 44.7 and 47.4 in 1891–1900; by 1950–2 the figures had climbed to 64.4 and 68.7, and in 1996–8 to 72.4 and 77.9 (note the widening absolute gender gap in the twentieth century).
But this was not the only widening gap. When comparable data first became available, for the 1870s, e0 for England and Wales was less than six months longer than in Scotland and about ten months longer for women. The situation in Scotland was better than France and Germany, but well behind Denmark, Sweden, and Norway. By 1930–2 the gap compared with England and Wales was 2.7 years for men and over 3.4 years for women (for women compared with Sweden it was over six years); these gaps only narrowed slowly over the rest of the century.
However, in one area there was marked relative improvement, at least until after the Second World War. As Chapter 24 shows, in great part due to their dramatically rapid growth, Scottish nineteenth-century cities had appalling death rates (for example, e0 in Glasgow even in the 1890s was almost ten years below the Scottish average). In the 1860s most rural areas had far lower death rates than small towns, and they in turn than those of the large towns and cities. Over the century up to 1950, marked improvements in sanitation and living conditions brought crude death rates down much faster in larger towns and cities than in the smaller towns and the countryside, though this was helped by the general ageing of rural populations. However, in the last quarter of the twentieth century, most of Greater Glasgow (and also the Western Isles) experienced slower improvement in mortality from key diseases than most of the rest of the country; as male death rates from heart disease fell dramatically for Scotland as a whole, the relative position of people living in many LAs in the west of Scotland worsened; the same was true of female lung-cancer deaths, which rose faster in these areas than in Scotland as a whole.
Figure 2.4 reflects the principal changes in overall mortality in Scotland from the 1860s to 2000–2. It shows the numbers of females born in any year who would have survived to any particular age had the mortality conditions of that year not changed during their lifetimes (for example, at 1860s mortality levels, roughly half of all girl babies would have survived to the age of 50, compared with around three-quarters at 1930–2 levels and over 95 per cent if 2000–2 rates continued). It also captures the relative impact of deaths at different ages on survival chances at any particular point in time. Glasgow in the 1890s is also shown for comparison. The figures for males were always rather worse but broadly moved in parallel.
Over the second half of the nineteenth century, infant death rates actually increased, with more than one in eight babies not reaching their first birthday by the 1890s; this, above all, kept expectation of life at birth so low. Thereafter there was steady improvement as family size fell and sanitation and baby-care improved, but even in the late 1930s infant mortality was still above 70 per 1,000 live births (about 100 in Glasgow, 90 in Dundee). The key change came after the Second World War, and by the end of the century infant mortality was down to about one in 200 births.
By contrast, big improvements from improved sanitation and living standards impacted much earlier on deaths, especially from infectious diseases, of children and young adults. In the 1860s a fifth of baby girls died between their first and fifth birthdays, by the 1890s this was down to one in eight, and by the 1930s to one in eighteen; aided by antibiotics, by 2000–2 it was one in 200. And of women reaching 15 but dying before 45, the parallel figures were a quarter, a fifth, one in nine, and, once mortality associated with childbirth and tuberculosis had been brought under control after the 1930s, one in fifty.
Improvements were much slower at older ages, where most deaths were always from chronic conditions—indeed, increasing tobacco consumption with a consequential marked rise in respiratory cancers was a key factor ensuring that expectation of life for men aged 60 only improved by one year between the 1860s and 1950–2, when it was 14.3 years. It reached 15.3 in 1980–2, and then, as tobacco smoking, cholesterol, and high blood pressure all became more controlled, it rose much faster to 18.5 by 2000–2.
FIGURE 2.4 Survivors per 100,000 females born, Scotland and Glasgow, various dates.
What were the impacts of the combination of fertility, mortality, and migration changes on the age structure of the Scottish population? In 1851, 36 per cent of Scots were aged under 15 and just 4.8 per cent were 65 or over. At censuses from 1911 more detail is available: see Figure 2.5.
The graph for 1911 reflects a society where population had been growing rapidly, with more people born in each decade than in its predecessor, but where relatively high mortality at all ages pulled down the numbers surviving to older age groups. As a result, the proportion of the population aged 65 and over in 1911 was only 5.4 per cent, while the proportion under 15 was 32 per cent.
FIGURE 2.5 Age structures of the Scottish female population, 1911, 1951, and 2001.
By 1951, the pattern had changed. Low fertility in both World Wars was followed by post-war baby booms, but, above all, the graph reflects low interwar fertility and the fertility decline and high emigration before 1914. Fewer than 25 per cent of Scots were under 15. Survival at older ages was still restrained, leaving 9.9 per cent of the population aged 65 and over.
The 2001 pattern changed again. The birth-rate boom of the 1960s and the subsequent erratic decline are well shown: just 18 per cent were aged under 15. But rapidly extending survival chances of the older population meant that 16 per cent were aged 65 and over.
Note finally, however, that even if everyone aged under 20 and over 64 is defined as dependent, this still produces a dependent percentage of 40.1 per cent for 2001, below the figure for 1951 and more than ten percentage points below 1851 (when many people even aged 65 were far less capable of work than at the start of the twenty-first century).
The years 2001–10 saw growing media and political concern over the present and future state of Scotland’s population. One contributor was undoubtedly the more accessible information emerging from the General Register Office when the more discursive and accessible Annual Reviews replaced the old Annual Reports, and a growing range of information was regularly publicized on the GRO(S) website.37 A second factor was a growing interest by devolved government in population matters, starting with the initiative on migration policy in 2004 and accelerating with the SNP administration’s population objectives in 2007. Third, there was, at long last, a resurgence of academic interest in population issues, enthusiastically supported personally by the Registrar General and with funding by the Economic and Social Research Council.
These developments paralleled two marked changes in demographic trends—though only time will tell whether either will survive the post-2008 recession and the 2010 Westminster government’s pledge to reduce UK immigration. The change that drew most comment was a marked reversal of the downward trend in national population, which had fallen from a 1971 peak of 5.24 million to a low point of 5.05 million in 2002, on trend to below five million by 2009. Instead it recovered, to an estimated 5.19 million in 2009. This was not the first reversal in three decades of decline (population had actually risen in every year between 1989 and 1995); but this time the average rate of growth for 2003–9 was consistently higher than in any comparable period since the early 1930s.
Positive net migration, a second ‘new’ feature, was also not new (compare four years in the early 1990s and 1931–2), but it, too, at an annual average of more than 20,000, was the highest figure ever recorded for a period of this length. What was also new was that this positive net migration came mostly, not from falling outflow as in the past, but from increased inflow, especially from the new EU accession states. These immigrants also started to have children and this contributed to another reversed trend, a modest rise in fertility.
But, alongside these new developments, which are still too new for us to judge whether they are more than temporary blips, many of the old patterns and trends continued. While Glasgow City’s population showed a small increase for the first time in decades (largely a function of new immigrants), the shifting balance of population from the west Central Belt towards the east continued, with fastest rates of decline for 1999–2009 in Inverclyde and the Western Isles, fastest growth rates in West and East Lothian, and highest absolute growth in the City of Edinburgh. Over the same period, age-standardized death rates for Scotland also fell, with continuing reduced mortality from heart and circulatory conditions and respiratory cancers for men; but lung-cancer rates for women rose further, and alcohol-related deaths for both sexes in 2009 were two and a half times their 1989 level. Equally depressingly, between 1996–8 and 2006–8, the gap between the local authorities with lowest and highest expectation of life at birth remained constant for men and actually increased for women. Also, between 1999 and 2009: the mean age at marriage continued to rise (by over two years for both men and women); the number of marriages (after a modest temporary recovery) continued to fall but with no evidence that rates of cohabitation were declining; the proportion of births to non-married women rose by more than a fifth, but the proportion registered solely in the mother’s name fell further, implying that much of the increase was to mothers living in a relationship of some kind. But, overall, we still understand so little about the deeper causes of any of the current trends, that guesses as to future trends or the wider implications for Scottish society seem hopelessly premature.
Anderson, Michael, ‘Population and Family Life’, in Tony Dickson and James Treble, People and Society in Scotland 1914–1990 (Edinburgh, 1992), 12–47.
—— and Morse, Donald, ‘The People,’ in Hamish Fraser and R. J. Morris (eds.), People and Society in Scotland 1830–1914 (Edinburgh, 1990), 8–45.
Flinn, Michael et al., Scottish Population History from the 17th Century to the 1930s (Cambridge, 1977).
Houston, R. A., ‘The Demographic Regime’, in T. M. Devine and Rosalind Mitchison, People and Society in Scotland, 1760–1830 (Edinburgh, 1988), 9–26.
Rothenbacher, Franz, The European Population, 2 vols.: 1850–1945 and Since 1945 (Basingstoke, 2002 and 2005).
Scotland’s Population 2004: The Registrar General’s Annual Review of Demographic Trends, 150th Edition (Edinburgh, 2005).
Time Series Datasets (Edinburgh: General Register Office for Scotland), http://www.gro-scotland.gov.uk/statistics/time-series-datasets.html, updated annually.
Tyson, Robert, ‘Demographic Change’, in T. M. Devine and J. R. Young, Eighteenth-Century Scotland: New Perspectives (East Linton, 1999), 195–209.
Whyte, Ian, Migration and Society in Britain, 1550–1830 (Basingstoke, 2000).
