by Paul Morland
Over the course of this modern period, as conflict is decreasingly between states and increasingly between ethnic groups, often within states, demography becomes important precisely because contending ethnic groups often have starkly different demographic profiles.18 Given the importance of numbers to ethnic and national groups in conflict, one might expect such groups to adopt strategies which aim to advance their own demographic strength, either through increasing their own or decreasing those of rivals or both. These strategies, known collectively as demographic engineering, can be of the ‘hard’ or ‘soft’ variety. Hard demographic engineering involves the creation, destruction and movement of people through policies such as selective incentives to fertility, genocide or the encouragement of inward or outward flows of people from a given territory. Examples are sadly numerous. In the 1920s the United States explicitly shaped its immigration policy to preserve its ‘Anglo-Saxon’ character against further incursions from southern and eastern Europe. In the middle years of the twentieth century, Protestant leaders in Northern Ireland tacitly encouraged Catholic emigration while Catholics adopted higher birth rates in part to boost their numbers. Sri Lanka’s Sinhalese-dominated government expatriated Tamils of relatively recent South Indian origin to boost the Sinhalese nature of the state. In Communist Romania, ethnic Hungarians were given readier access to contraceptives and abortions than ethnic Romanians while ethnic Germans and Jews were encouraged to leave the country altogether, all in the name of boosting the ethnic Romanian character of the country.19
Soft demographic engineering, by contrast, while still concerned with demographic policies aimed at enhancing the numbers of one group over another, uses means such as redrawing frontiers, manipulating identities or manipulating censuses and census categorisations. Examples include the consolidation of the Sinhalese identity in Sri Lanka out of Kandyan highlanders and people of the Low Country, or the suggestion that Turkey redefine Kurds as ‘mountain Turks’.20 This is one way in which demography is shaping destinies.
It is reckoned that whereas in the 1950s around half the world’s conflicts were between states and half within them, in the 1990s the latter outnumbered the former by a ratio of six to one. Whereas 57% of conflict during 1945–2008 was ‘ethnic’, all conflict between 2000 and 2010 can be given that label.21 Christians, once dominant in Lebanon, have been eclipsed by Muslims, who sustained a higher birth rate over a longer time and were less likely to leave the country. Today the main struggle for power in Lebanon is between Sunni and Shi’a Muslims rather than between Muslims and Christians. Whether in civil wars, referendums or elections, numbers make the difference between dominance and marginalisation, victory and defeat, at home as well as abroad. When some groups have sky-high birth rates or arrive in serried ranks while others have tiny families or up-sticks, it is usually demography that determines who controls communities, regions and countries.
It is worth spelling out at this point that nations and ethnic groups are real and that they matter in history. Humans are not a naturally solitary species, but live in groups. Loyalties are initially to the band or tribe. A shared sense of common ancestry, language and custom is universal in hunter-gatherer societies. How these sentiments are transformed in complex and modern societies is subject to much scholarly debate, but the fact that they exist is undeniable. These affiliations explain much of how the world works, and has worked in the recent past, including the outcome of conflicts and elections.
It is certainly true that many people would like to overlook the highly ethnic nature of politics around the world and assume that our cosmopolitan preferences will become more universal if we ignore the nationalism and ethnocentrism of others. However, in much of the world, ethnicity matters politically. And pretty much everywhere, it has mattered at least until recently. It may be that some genuinely post-ethnic, multicultural societies are beginning to emerge in some of the more urban and cosmopolitan parts of the West (coastal USA and London, for example), but even within those places there are populist backlashes. Both Brexit and Trump can be understood as part of that backlash.
A Brief Guide to Demography
To understand how demography has driven history it is necessary first to outline its three timeless fundamentals. The good news is that this is fairly simple. Only three things can change the number of people in a region or country: the first is births, which provide additions to the population; the second is deaths, which cause subtractions from it; and the third is migration, or the net movement of people in or out.
The birth rate (sometimes called the ‘crude birth rate’) is simply the number of births relative to the population. The death rate (known as the ‘crude death rate’, or ‘crude mortality rate’) is the number of deaths relative to the population. For example, in England and Wales in 2014 there were around 700,000 births in a population of 58 million, giving a crude birth rate of about twelve per thousand.22 (Note that demographic data is often presented as ‘per thousand’ rather than per hundred or per cent.) In the same year, England and Wales had around half a million deaths, or a crude death rate of about eight and a half per thousand. Without any immigration or emigration, this would have given England and Wales a population growth rate of 3.5 (i.e. 12–8.5) per thousand, or 0.35%. This is equivalent to about 200,000 people, the gap between births and deaths. In the United States, the figures for the crude birth rate are about twelve and a half per thousand and the mortality rate is just over eight per thousand, which results in annual population growth, excluding immigration, of almost 1.5 million a year. In Germany, which has experienced years of declining population, the crude birth rate is around eight per thousand, whereas the crude mortality rate is just under eleven per thousand. Without immigration, the German population would be falling by nearly a quarter of a million a year.
In many developing countries, particularly (but not exclusively) in Africa, birth rates are very high while death rates have fallen considerably. Even very basic health care and nutrition improve infant mortality, extend life expectancy and so decrease death rates quite substantially. Sub-Saharan Africa as a whole has a crude birth rate of around thirty-eight per thousand, compared to Europe’s modest eleven. In the middle of the twentieth century, Africa’s crude death rate was not much short of thirty; today, it is not much more than ten. Both Iraq and Afghanistan have high birth rates (around thirty-five per thousand) and both, despite all the violence they have suffered, have managed to bring down their death rates; between the late 1990s and 2010–15 the death rate in Afghanistan fell from above thirteen per thousand to below eight, in Iraq from an already low 5.7 to 5.3. Most people would be surprised to learn that Iraq’s death rate is lower than the UK’s. It is testimony to the youthfulness of Iraq’s population and to the fact that, as in Afghanistan, while the violence on our television screens concerns the deaths of tens or even hundreds of thousands, improvements in nutrition and health care affect tens of millions. This is why, even in the second decade of the twentieth century when Europe was plunged into the First World War, followed by a deadly influenza epidemic, the continent’s population continued to grow.
The advantage of crude birth and death rates is that they are simple and they tell us how quickly a population is growing or declining. Their shortcoming–the reason why birth and death rates are often called ‘crude’–is that they do not take account of the age structure of a country. You would expect to see more deaths relative to the population in a country like Japan, which is full of old people, than in Ireland, which is still relatively young. Likewise, you would expect there to be more births per person in Ireland where there are, relative to the population, more women of childbearing age, than in Japan. To adjust for this, demographers also measure total fertility rate and life expectancy. These indicators describe how many children the average woman can expect to have–regardless of how many young women there happen to be in a given population–and how long the average person can expect to live–regardless of how old the general populat
ion is. (‘Fertility’, therefore, means actual childbearing rather than the biological potential of bearing children. A woman who is perfectly fertile, that is, capable of having a child or many children, may for a variety of reasons never have any. When demographers speak of fertility they refer to children actually born.) These expectations are based on the actual births to women and the actual deaths of people at different ages. For more on this, see the Appendices at the end of this book.
Fertility rates are always quoted ‘per woman’ for a number of reasons. First, there is near certainty about who is the mother in each case of a birth; the father’s identity is more uncertain. Counting births per father might mean double counting or leaving some births out. Second, the number of children a woman can have ranges from zero to very rarely more than fifteen. For a man it ranges from zero to (at least in theory) thousands, so the fertility rate as a number is more easily manageable for women. Third, there is greater certainty about the cohort of women likely to have children than there is about the cohort of men. Statistically, the fertility for women beyond the age of around forty-five can more or less be ignored. Older women do have children, but rarely enough for it to impact the statistics meaningfully. Men, by contrast, can in theory at least continue having children to the end of their lives. So demography invariably focuses on women, at least when it comes to births, although in doing so it sometimes has a tendency to view them as statistics or units, interesting to the extent to which they produce children or not. Whilst childbearing can and indeed must be viewed statistically, comparing it between places and between times and seeing how it changes, far more insight may be gained by viewing these changes through the lives and choices of individual women, giving voice to their aspirations, anxieties and decisions. This is not just about illustrating the data but also about showing how one of the most inspiring elements to the story of population in the last two hundred years is how women have progressively been able to take control over their own decisions and bodies.
The difference between crude birth rate and fertility rate may be illustrated by comparing South Africa and Israel. South Africa has witnessed a rise in women’s education, urbanisation and concerted government efforts to provide birth control services, and as a result has experienced a sharp fall in fertility–in this it is well ahead of the rest of sub-Saharan Africa. But because, until fairly recently, fertility rates were high, there are many young people as a share of the total population, a reflection of the fertility choices of an earlier generation. Israel, in contrast, is an unusual case: a developed country that has actually seen the number of children born per woman rise in recent decades. South Africa has a slightly higher birth rate than Israel: twenty-two per thousand compared to twenty-one. But this is not caused by the average South African woman having more babies. Rather, there are simply more young women who are having babies. In Israel, a woman on average has more than three children now, versus less than 2.5 in South Africa. South Africa’s marginally higher birth rate is a product of its recently (but no longer) high fertility rate; as recently as the late 1970s it was five children per woman. This has generated a young population, full of women of childbearing age, but they are now not choosing to have many children. In Israel, by contrast, the total fertility rate in the late 1970s was a child and a half lower than the South African level. There are fewer women of childbearing age as a proportion of the overall population in Israel, but they are each having more children, and so Israel has a high fertility rate (the average number of children born to an individual woman) but not a particularly high crude birth rate (the number of births relative to the population as a whole).
Total fertility rate is a good measure of the moment, a snapshot of what is happening to fertility at a particular point in time. A definitive measure is completed fertility for a cohort or generation, but this is only available, as it were, after the event, that is, when all of the women of that cohort have passed their fertile years. It is possible to compare the number of children German women born in the 1870s had to those born in the 1890s; none of them will be having any more. But it is more difficult to compare definitively the childbearing of German women born in the 1970s and the 1990s; both groups may still not have completed their fertility and may still have childbearing ahead of them. Total fertility is the best measure available for determining what is happening to fertility now.
Like mortality rates, birth rates give us a crude measure for the population as a whole; like life expectancy, fertility rates give us a measure tailored to the specific structure of the population. Take Japan and Guinea, in west Africa. Both countries have a crude mortality rate of ten per thousand. But the reasons for this similar level of mortality could not be more different. Japan is an old country; Guinea is a very young one. If they had the same life expectancy, Guinea would have a much lower mortality rate than Japan because there would be far fewer deaths among its population, which is young, than among Japan’s, which is old. As the crude mortality rates of these countries are the same, and Guineans are on average much younger than the Japanese, then Guineans must be dying much younger than people in Japan. People in Japan can expect to live into their mid eighties. For most Guineans, life ends nearly thirty years earlier. Think of a boarding school and an old people’s home, each with a thousand inhabitants. If twenty people died in both institutions in a given year, they would both have a crude death rate of twenty per thousand, but people in the elderly residential care home on average would be living much longer lives than people in the school.
The age structure of a population can also be analysed by showing what percentage are small children or are over sixty-five. The simplest way is to figure the median age: if all the population was lined up in age order, how old would the person in the middle be? In Guinea, the median age is below nineteen; in Japan it is over forty-six.
The Data
Understanding and tracking all of this depends on data. The data is certainly not uniform, nor is it uniformly accurate over time and space. Generally, the later the data and the more developed the country, the greater its reliability.23 British censuses, which started in the early nineteenth century, are generally reliable in terms of overall population size. The birth of the life insurance industry–whereby insurers needed to calculate the probability of someone at a given age dying–means that we have a good idea of the mortality rates and life expectancies in some places dating back to the eighteenth century. In some places, local records, usually those of the parish, have been expertly extrapolated by demographers to build a picture of wider society. In other countries, censuses go back a long way. Indeed, the census is almost as old as the state, and testifies to an inherent desire of states to know about their inhabitants, sometimes predominantly for military reasons, at other times for tax purposes. There were censuses in ancient Egypt and China thousands of years ago, while the Bible speaks of censuses in ancient Israel–like the Chinese, the Romans had censuses mostly to determine military capacity. The Ottomans had censuses for the same reason, but therein lies the flaw, at least for the purposes of historians. Only Muslim men of a certain age were qualified to fight in Ottoman armies, so only these were of interest to the census officers. To figure out the size of the entire population at various times, extrapolations and assumptions are required. Censuses continue in most developed countries today, and recent proposals to end them–suggesting alternative sources of data were now available, that sampling would suffice given sophisticated statistical techniques, and that costs could be saved–have been vigorously and successfully opposed.
Today, demographers are blessed with a standard set of measures from a variety of sources, not least the United Nations, which offers records of birth and death rates, fertility, longevity and median age by country and continent, going back to 1950 and with projections to the end of the twenty-first century. No data is perfect, but that of the United Nations is recognised as being of high quality, and so it is heavily relied on for the later chapters of this book. Wh
ere the reliability and quality of data is open to question, I have used the best available sources but have not opened up the topic to lengthy discussion.
To get a sense of what the data means, Table 1 provides a useful guide to what is high, what is low and what is experienced today in the UK.
Table 1: Demographic Data: What’s High, What’s Low
What makes demography exciting, what makes it a more powerful tool for understanding the world than is generally appreciated, is that each of these numbers can be seen in three ways. First, in and of itself, as an illustration of something meaningful about a society. For example, the fact that the UAE has such an extraordinarily low mortality rate is testimony both to the huge recent growth of its population (there are few elderly people relative to the population: 2% of people in the UAE are over sixty compared to 12% worldwide and 27% in Germany), the extraordinarily long life expectancy (only a couple of years short of the USA’s) thanks to world-class health care and public health and the enormous size of the immigrant population (90% of the total), most of whom are likely to go home to south Asia or Europe and not die in the UAE. By unpacking one piece of data, a great deal of light is shone on today’s UAE.
Second, taken as part of a chain, it can illustrate extraordinary change. Kenya’s population may have been growing at nearly 4% annually in 1982 but by the year 2000 the growth rate had fallen to 2.5% thanks to a success in reducing the fertility rate from seven to five. (Since then the growth rate has plateaued, meaning that Kenya’s population continues to grow fast although somewhat less quickly than in the 1980s.)