It is not easy to make historical demography readable. Perhaps because of his unrivalled knowledge of Scottish population history Michael Anderson also has difficulties. It is as if the sociologist, the historian and the geographer in Professor Anderson were in conflict as he grappled with this difficult subject. The book is richly filled with detail and explanation of the population and economy of the varied localities of Scotland but the reader has to work hard to excavate them. The topic warrants a full-length review article but in this short assessment emphasis is on what Anderson tells us about Scotland's urban history.

There are 20 chapters organized into 6 sections: ‘Questions and contexts’, ‘Multiple Scotlands’, migration, fertility and nuptiality, mortality and conclusions. The context is geography and economy. Chapter 3 on the constraints of physical geography, if elementary, will be useful for those unfamiliar with the geography of Scotland. In the wider European economic context, it is imperative to remind readers that early nineteenth-century Scotland was one of the most industrialized regions in the world, yet between 1851 and 2001 the primary sector fell from 30.4 per cent to 3.6 per cent while services rose from 15.9 per cent to 53.6 per cent (Table 3.1). From the 1970s, employment in manufacturing fell from about one third to 13 per cent by the end of the century. A comparison with England and Wales would have been useful. In the urban context, four major cities (Glasgow, Edinburgh, Dundee and Aberdeen) contained 20 per cent of the population in 1851, more than 33 per cent in 1901 and nearly 40 per cent in 1951 (p. 35). Rightly, in relation to its population history, Anderson summarizes the role of the character and quality (or lack of it) in Scottish housing and sanitary arrangements, and the twentieth-century flight from tenements to suburbs and public housing. Living standards, the poor law, unemployment and a catch-all of ‘cultural factors’ are all summarized to leave the reader in no doubt that the causes of Scottish population were many and complex.

Part II, ‘Multiple Scotlands’, has four short chapters on the geography of Scottish population change. A series of seven maps (of which more later) presents population density and change at parish level from 1801 to 2001. The diversity of Scottish population change outside major urban areas and the islands is explored in detail but a clearer picture (not just hints) of constraints and possibilities of the physical environmental and resources would help the reader. Chapter 7, ‘The major urban centres’, will be of most interest to readers of this journal. We are told that by the 1860s, in common with ‘most towns in England’ (the European perspective being briefly forgotten), ‘Scottish towns had already moved to positive natural increase’ (p. 94). That is crucial and its comparative and chronological context should be at the centre of the argument. The old question of whether immigrants suffered higher mortality than those they left behind is raised but the comment that as yet we do not know the answer is frustrating. In a way, of course (and the author clearly understands all such snags), it is a chimera. Does not an immigrant in a town become a local? How many migrants were just passing through and for how long? At this stage, the reader is promised further discussion later but after reading the last three paragraphs on page 95 (too long to quote here) and learning that no two Scottish ‘towns’ are the same, an urban historian might want to turn to a more tractable topic. A rather arid discussion of what constitutes a town inevitably runs into the sand.

Emigration is highly significant in Scotland's history. Figure 8.2 leaves us in no doubt. Demographers have tended to regard migration as a nuisance. Anderson considers it before moving on to the natural components of population change, but after struggling through his painstaking analysis, a reader might conclude that migration is indeed a nuisance. Anderson tackles all the issues but it is hard to distinguish the firm ground from the quicksand and too often he reverts to his default mode of covering all his bases: ‘Differential patterns of internal migration played a key role in regional and local differences in population growth and decline; indeed for long periods they were more important than the differences in fertility and mortality (large though these sometimes were)’ (p. 136).

The longest section of the book is on marriage and nuptiality. Here, Anderson is at his best and, one suspects, most comfortable. It is certainly the most demographically significant part of the book. It could have been a book on its own though to appreciate it fully requires a knowledge of the methods and language of demography, not least the Princeton Fertility Project.

The chapters on mortality cover the national picture and spatial variations comprehensively. Given the size of its contribution to mortality in general, infant mortality might have received greater emphasis. The surprisingly low infant mortality rate was not just in rural areas but also in large cities. From 1876 to 1902, the nine largest cities in Scotland had a mean infant mortality rate of 143 per 1,000 live births. The largest provincial conurbations in England over the same period averaged about 180. Why?

Anderson tells us that ‘one of the principal arguments of this book is that it is usually highly misleading to treat Scotland as a single coherent socio-spatial unit’ (p. 13). That strikes a population geographer as a statement of the obvious rather than an argument. But here also lies an unresolved tension: between this and his detailed presentation of Scotland distinct from England and Wales and other European countries, Anderson comes close to saying that the regional and local variations are not that significant after all (p. 428).

It is a disgrace that this book, by a leading Scottish academic, university vice principal and chairman of the National Library, was not published by a Scottish university press. Sad to say, however, the author has also been ill-served by Oxford University Press. Instead, we have an expensive monograph with diagrams that are frankly dreadful. Many of the graphs are reminiscent of the early output from Microsoft Excel; some are illegible and others could have been replaced by well-chosen words. The 13 valuable maps are in greyscale; in colour, they alone would sell the book. In 2018, there was no excuse for this. Compare the coloured maps in Robert Woods’ The Demography of Victorian England and Wales (Cambridge University Press, 2000). It is unfair that this should reflect on a distinguished author. In some ways, it is a disorderly book, perhaps because the author has not been clear for whom it is written, but Oxford University Press should have higher editorial standards to benefit authors as well as readers.