Clark (2007).pdf

 The Logic of the Malthusian Economy No arts; no letters; no society; and which is worst of all, continual fear, and danger of violent death: and the life of man, solitary, poor, nasty, brutish and short. —Thomas Hobbes (1651)1 The vast majority of human societies, from the original foragers of t...

 The Logic of the Malthusian Economy No arts; no letters; no society; and which is worst of all, continual fear, and danger of violent death: and the life of man, solitary, poor, nasty, brutish and short. —Thomas Hobbes (1651)1 The vast majority of human societies, from the original foragers of the African savannah through settled agrarian societies until about 1800, led an economic life shaped and governed by one simple fact: in the long run births had to equal deaths. Since this same logic governs all animal species, until 1800 in this “natural” economy the economic laws for humans were the same as for all animal species. The break between the economics of humans and the economics of the rest of the animal world occurred within the past two hundred years. It is commonly assumed that the huge changes in the technology available to people and in the organizational complexity of societies, between our ancestors on the savannah and those in England at the time of the Industrial Revolution, must have improved material life even before modern economic growth began. For example, Angus Maddison, the much-quoted creator of preindustrial economic data, hazarded estimates of income per person for millennia before 1820 on this basis.2 But in this chapter I show that the logic of the natural economy implies that the material living standard of the average person in the agrarian economies of 1800 was, if anything, worse than that of our remote ancestors. Hobbes, in the quote that opens this chapter, was 1. Hobbes, 1651, 84. 2. Maddison, 2001, 28, for example, estimates that GDP per capita in western Europe more than doubled from $450 in AD 1 to $1,232 by 1820 (in 1990 dollars), while for Japan the rise was from $400 to $669.  profoundly wrong to believe that man was any worse off in the natural state than in the England of 1651. This chapter develops a model of the preindustrial economy, the Malthusian model, from three simple and seemingly innocuous assumptions. This model has profound implications for how the economy functioned before 1800, which are then tested and explored in the following four chapters. The Malthusian Equilibrium Women, over the course of their reproductive lives, can give birth to twelve or more children. In some current societies the average woman still gives birth to more than six children. Yet in the world before 1800 the number of children per woman that survived to adulthood was always just a little above two. World population grew from perhaps 0.1 million in 130,000 BC to 770 million by 1800. But this still represents an average of 2.005 surviving children per woman before 1800. Even within successful preindustrial economies, such as those in western Europe, long-run rates of population growth were very small. Table 2.1 shows population in 1300 and 1800, and the implied numbers of surviving children per woman, for several western European countries. None of these societies deviated far from two surviving children per woman. Some force must have kept population growth rates within rather strict limits over the long run. The Malthusian model supplies a mechanism to explain this long-run population stability. In the simplest version there are just three assumptions: 1. Each society has a birth rate, determined by customs regulating fertility, but increasing with material living standards. 2. The death rate in each society declines as living standards increase. 3. Material living standards decline as population increases. The birth rate is just the number of births per year per person, for convenience normally quoted as births per thousand people. Maximum observed fertility levels have been 50–60. But the birth rate varies significantly even across preindustrial societies. Preindustrial England sometimes had birth rates of less than 30. As recently as 2000 in Africa, the area of highest birth rates, some countries had rates exceeding 50 per thousand: Niger, 55; Somalia, 52; Uganda, 51.    Table . Populations in Western Europe, 1300 and 1800 Location Norwaya Southern Italyb Francec Englandd Northern Italyb Icelanda Population ca. 1300 Population ca. 1800 Surviving children per woman 0.40 4.75 17.0 5.8 7.75 0.084 0.88 7.9 27.2 8.7 10.2 0.047 2.095 2.061 2.056 2.049 2.033 1.930 Sources: aTomasson, 1977, 406. bFederico and Malanima, 2004, table 4. cLe Roy Ladurie, 1981, 13.dClark, 2007a, 120. The death rate is again just deaths per year per person, also typically quoted per thousand people. In a stationary population, one of constant size, life expectancy at birth is the inverse of the death rate.3 Thus if death rates are 33 per thousand, life expectancy at birth is thirty years. At a death rate of 20 per thousand, life expectancy would rise to fifty. In a stationary population birth rates equal death rates. So in stationary populations, which were characteristic of the preindustrial world, life expectancy at birth is also the inverse of the birth rate. Thus in preindustrial society the only way to achieve high life expectancies was by limiting births. If preindustrial populations had displayed the fertility levels of modern Niger, life expectancy at birth would have been less than twenty. The material living standard refers to the average amount of goods and services (e.g., religious ceremonies, barbers, servants) that people in a society consume. When new goods are introduced over time, such as newspapers, Wedgwood fine porcelain, and vacations at the seaside, it can be tricky to compare societies in terms of the purchasing power of their real wages. But for most of human history, and for all societies before 1800, the bulk of material consumption was food, shelter, and clothing, so their material living standards can be measured more accurately. In societies sophisticated enough to have a labor market, the material living standard for the bulk of the 3. Formally, if e0 is life expectancy at birth and D is the death rate, e0= 1/D.                               Birth rate, death rate Birth rate Death rate Population y* y0 N * N0 y* y0 Income per person Figure . Long-run equilibrium in the Malthusian economy. population will be determined by the purchasing power of the wages of unskilled workers. Figure 2.1 shows graphically the three assumptions of the simple Malthusian model.4 The horizontal axis for both panels is material income, the amount of goods and services available to each person. In the top panel birth and death rates are plotted on the vertical axis. The material income at which birth rates equal death rates is called the subsistence income, denoted in the figure as y*. This is the income that just allows the population to reproduce itself. At material incomes above this the birth rate exceeds the death rate and population is growing. At material incomes below this the death rate exceeds the birth rate and population declines. Notice that this subsistence income is determined without any reference to the production technology of the society. It depends only on the factors that determine the birth rate and those that 4. The graphical exposition here follows that of Lee and Schofield, 1981.    determine the death rate. Once we know these we can determine the subsistence income and life expectancy at birth. In the bottom panel population is shown on the vertical axis. Once we know population, that determines income and in turn the birth rate and death rates. With just these assumptions it is easy to show that the economy will always move in the long run to the level of real incomes at which birth rates equal death rates. Suppose population starts at an arbitrary initial population, N0 in the diagram. This will imply an initial income y0. Since y0 exceeds the subsistence income, births exceed deaths and population grows. As it grows, income declines. As long as the income exceeds the subsistence level population growth will continue, and income will continue to fall. Only when income has fallen to the subsistence level will population growth cease at the equilibrium level, N*, and the population stabilize. Suppose that instead the initial population had been so large that the income was below subsistence. Then deaths would exceed births and population would fall. This would push up incomes. The process would continue until income was again at the subsistence level. Thus wherever population starts from in this society it always ends up at N*, with income at subsistence. The term subsistence income can lead to the incorrect notion that in a Malthusian economy people are all living on the brink of starvation, like the inmates of some particularly nasty Soviet-era gulag. In fact in almost all Malthusian economies the subsistence income considerably exceeded the income required to allow the population to feed itself from day to day. Differences in the location of the mortality and fertility schedules across societies also generated very different subsistence incomes. Subsistence for one society was extinction for others. Both 1400 and 1650, for example, fell within periods of population stability in England, hence periods in which by definition the income was at subsistence. But the wage of the poorest workers, unskilled agricultural laborers, was equivalent to about 9 pounds of wheat per day in 1650, compared to 18 pounds in 1400. Even the lower 1650 subsistence wage was well above the biologically determined minimum daily requirement of about 1,500 calories a day. A diet of a mere 2 pounds of wheat per day, supplying 2,400 calories per day, would keep a laborer alive and fit for work. Thus preindustrial societies, while they were subsistence economies, were not typically starvation economies. Indeed, with favorable conditions, they were at times wealthy, even by the standards of many modern societies.                               The assumption that is key to the observation that income always returns to the subsistence level is the third one, of a fixed trade-off between population and material income per person. For reasons given below, this tradeoff is called the technology schedule. The justification for the decline in material incomes with higher population is the famous Law of Diminishing Returns introduced to economics by David Ricardo (and independently by Malthus). Any production system employs a variety of inputs, the principal ones being land, labor, and capital. The Law of Diminishing Returns holds that, if one of the inputs to production is fixed, then employing more of any of the other inputs will increase output, but by progressively smaller increments. That is, the output per unit of the other input factors will decline as their use in production is expanded, as long as one input factor remains fixed. In the preindustrial era land was the key production factor that was inherently fixed in supply. This limited supply implied that average output per worker would fall as the labor supply increased in any society, as long as the technology of that society remained unchanged. Consequently average material income per person fell with population growth. Figure 2.2 shows the assumed relationship between labor input and the value of output for preindustrial societies that underlies the third assumption of the Malthusian model. In economics the increase in the value of output from adding one more worker is called the marginal product of that person. In market economies this equals the wage.5 As can be seen in the figure, the marginal product declines as more workers are added, and so does the wage. Average output per person also falls as the population rises. The additional output from the last person added to the economy is less than the output per person from existing workers.6 To appreciate concretely why this will happen, consider a peasant farmer with 50 acres of land. If he alone cultivates the land then he will maximize output by using low-intensity cultivation methods: keeping cattle or sheep which are left to fend for themselves and periodically culled for meat and hides, as with the Argentinean pampas in the early nineteenth century. With the labor 5. This is just the slope of the curve at any labor input. 6. Average output per person is the slope of the straight line drawn from the origin to the output curve at any given level of labor input.    Total output, marginal output Total output Marginal output Number of workers Figure . Labor input and output on a given area of land. of an additional person milk cows could also be kept, increasing total output. With yet more labor the property could be cultivated as arable land with grain crops. Arable land requires much more labor input per acre than pasture, given the need for plowing, sowing, harvesting, threshing, and manuring. But arable land also yields a greater value of output per acre. With even more people the land could be cultivated more intensively as garden land, growing vegetables and tubers as well, thus increasing output yet further. Yields are increased by ever more careful utilization of manure, and by suppression of competing weeds by manual hoeing. With enough labor input the output of any acre of land can be very high. In the agricultural systems of coastal China and Japan around 1800, an acre of land was enough to support a family. In Ireland before the potato famine of 1845, an acre of potatoes, with careful spade husbandry, could supply to a family more than 6 tons of potatoes a year, 36 pounds a day, nearly enough to subsist on.7 In the same period in England there were nearly 20 acres of land per farm worker. We can also see in figure 2.1 that the sole determinants of the subsistence income are the birth rate and death rate schedules. Knowing just these we can determine the subsistence income. The connection shown in the lower panel between income and population level serves only to determine the population that corresponds to the subsistence income. 7. Ó Gráda, 1999, 227.                               Birth rate, death rate Birth rate1 Birth rate0 Death rate Population y1* y0* N1* N0* y1* y0* Income per person Figure . Changes in the birth rate schedule. Changes in the Birth Rate and Death Rate Schedules Different societies will have different birth rate and death rate schedules, that is, the birth and death rates at given incomes, and these schedules can change within a society over time. Suppose, for example, that the birth rate schedule increased, as in figure 2.3. It is then simple to see what happens to the death rate, material incomes, and the population. In the short run births exceed deaths. Population thus grows, driving down real income, and increasing the death rate until deaths again equal births. At the new equilibrium real income is lower and population is greater. Any increase in birth rates in the Malthusian world drives down real incomes. Conversely anything which limits birth rates drives up real income. Since life expectancy at birth in the Malthusian era was just the inverse of the birth rate, as long as birth rates remained high, life expectancy had to be low. Preindustrial society could thus raise both material living standards and life expectancy by limiting births.    Birth rate, death rate Birth rate Death rate0 Death rate1 Population y1* y0* N1* N0* y1* y0* Income per person Figure . Changes in the death rate schedule. Again if the death rate schedule moves down, as in figure 2.4, so that at each income there is a lower death rate, then at the current income births exceed deaths, so that population falls. This again drives down real income until the death rate once more equals the birth rate. At the new equilibrium population is higher and income lower. Given the now lower birth rate, however, life expectancy would be somewhat higher. So improvements in sanitation, or declines in violence and disorder, which reduce the death rate schedule in preindustrial societies, can raise life expectancy, but only at the cost of lower material living standards. This Malthusian world thus exhibits a counterintuitive logic. Anything that raised the death rate schedule—war, disorder, disease, poor sanitary practices, or abandoning breast feeding—increased material living standards. Anything that reduced the death rate schedule—advances in medical technology, better personal hygiene, improved public sanitation, public provision for harvest failures, peace and order—reduced material living standards.                               Birth rate, death rate Birth rate Death rate Population y* N1* N0* y* Income per person Figure . Effects of isolated technological advance. Changes in Technology The real income in Malthusian economies was determined from the birth rate and death rate schedules alone. Once this income was determined the population size depended just on how many people could be supported at this income level given the land area and technology of the society. For each society, depending on the land area and the production technology, there was a schedule connecting each population level with a given real income level. This is called the technology schedule, because the major cause of changes in this schedule has been technological advance. But other factors could also shift this schedule: the availability of more capital, improved trade possibilities, climate changes, or better economic institutions. Figure 2.5 shows the path of adjustment from an isolated improvement in technology: a switch from an inferior technology, represented by curve T0, to a superior technology, represented by curve T1. Since population can change only slowly, the short-run effect of a technological improvement was an in-    crease in real incomes. But the increased income reduced the death rate, births exceeded deaths, and population grew. The growth of population only ended when income returned to subsistence. At the new equilibrium the only effect of the technological change was to increase the population. There was no lasting gain in living standards. The Malthusian Model and Economic Growth In the millennia leading up to 1800 there were significant improvements in production technologies, though these improvements happened slowly and sporadically. The technology of England in 1800—which included cheap iron and steel, cheap coal for energy, canals to transport goods, firearms, and sophisticated sailing ships—was hugely advanced compared to the technology of hunter-gatherers in the Paleolithic, before the development of settled agriculture. The degree of technological advance was revealed in the encounters between Europeans and isolated Polynesian islanders in the 1760s. The English sailors who arrived in Tahiti in 1767 on the Dolphin, for example, found a society with no metals. The Europeans’ iron was so valuable to the Tahitians that a single 3-inch nail could initially be bartered for a 20-pound pig or a sexual encounter. Given the enthusiasm of the sailors for the sex trade, nail prices two weeks later had dropped by half, and “the Carpenter came and told me every cleat in the ship was drawn, and all the Nails carried off . . . most of the hammock nails was drawn, and two-thirds of the men obliged to lie on the Deck for want of nails to hang their Hammocks.”8 When Captain Cook arrived at a similarly isolated Hawaii the local inhabitants on a number of occasions stole ship’s boats to burn them to retrieve the nails. But, though technology was advancing before 1800, the rate of advance was always slow relative to that in the world after 1800. Figure 2.6, for example, shows the actual location of the technology curve of the Malthusian model for England from 1200 to 1800. The figure shows income per person by decade versus population. The observations for each decade are linked to show the movement of the population and income combinations over time. English population showed dramatic variation in the preindustrial period. 8. Robertson, 1955, 32, 78, 104. When Captain Cook arrived in 1769 he was shocked to find that the locals now demanded a hatchet for a pig; Banks, 1962, 252.                               1790s Population (millions) 8 6 1310s 1600 4 1200s 1450s 2 0 0 20 40 60 80 100 Income per person (1860s = 100) Figure . Revealed technological advance in England, 1200–1800. There was growth in the medieval period from 1200 to 1316; at six million, the population in 1316 was as great as in the early eighteenth century. But the arrival from Asia of the bubonic plague (the so-called Black Death) in 1348 caused a long period of population decline from then to the 1450s.9 By then England had barely two million people. Population grew again from 1540 to 1640 as the plague loosened its hold. From 1200 to 1650, as population changed under the influence of disease shocks, the income-population points lie along one downward-sloping line. This implies a completely stagnant production technology for 450 years. After 1650 the implied technology curve shifts upward, but not fast enough to cause significant increases in output per person. Instead technological advance, as predicted, resulted mainly in more people. In particular in the later eighteenth century all technological advance created only a larger population without generating any income gains. Before 1800 the rate of technological advance in all economies was so low that incomes could not escape the Malthusian equilibrium. Because I want to show that the same economic model applies to all societies before 1800, even those with no labor market, and also to animal popula9. The popular term Black Death for the plague was introduced in England only hundreds of year after its onset.    tions, the Malthusian model has been developed in terms of material consumption per person. However, Thomas Malthus (1766–1834) and David Ricardo (1772–1823), who first formulated the Malthusian model and the associated economic doctrines now called classical economics, thought in terms of the wages of unskilled workers.10 Thus Ricardo, using similar logic, argued that real wages (as opposed to income per person, which includes land rents and returns on capital) must always eventually return to the subsistence level.11 Ricardo’s proposition later became known as the Iron Law of Wages. Classical economics thus denied the possibility of other than transitory improvements in the living standards of unskilled workers. All the above reasoning about birth rates, death rates, population, and incomes can be carried out equivalently in terms of wages. In light of subsequent events, the Iron Law of Wages may seem like an absurd proposition on which to found classical economics. But we shall see that the Malthusian model is an accurate description of all societies before 1800. The propositions of classical economics were developed at a time, 1798– 1817, when real wages in England had been stationary or declining for generations. Though the innovations associated with the Industrial Revolution began appearing in the 1760s, their significance was not widely appreciated at the time. Technical progress in production technologies still seemed modest, sporadic, and accidental before 1820. Figure 2.7 shows the rural setting of Malthus’s employment while working on his famous essay. Real wages did not begin the almost continual rise that characterizes the successful economies of the modern world until the 1820s. For some groups—such as the agricultural laborers in the south of England to whom Malthus ministered to as a parson while writing his Essay on the Principle of Population—real wages declined substantially between 1760 and 1820. Indeed one of the great social concerns of the years 1780–1834 in England was the rising tax burden on rural property owners created by payments to support the poor under the Poor Law. Thus Malthus and Ricardo predicted that, as long as fertility remained unchanged, economic growth could not in the long run improve the human condition. All that growth would produce would be a larger population living at the subsistence income. China, for Malthus, was the embodiment of this type of economy. Though the Chinese had made great advances in agricul10. They did so in part because in the era in which they wrote there were scant available measures of income per person. 11. McCulloch, 1881, 50–58.                               Figure . The church in Okewood, where Malthus earned his living as a curate while working on his essay. Malthus probably lived at his father’s house in nearby Albury, whose population of 510 in 1801 had grown to 929 by 1831. tural drainage and flood control, and had achieved high levels of output per acre, they still had very low material living standards because of the country’s dense population. Thus, Malthus wrote of China, “If the accounts we have of it are to be trusted, the lower classes of people are in the habit of living almost upon the smallest possible quantity of food and are glad to get any putrid offals that European labourers would rather starve than eat.”12 In the preindustrial world sporadic technological advance produced people, not wealth. Human and Animal Economies The economic laws we have derived in this chapter for the preindustrial human economy are precisely those that apply to all animal, and indeed plant, 12. Malthus, 1798, 115.    populations. Before 1800 there was no fundamental distinction between the economies of humans and those of other animal and plant species. This was also a point Malthus appreciated: “Elevated as man is above all other animals by his intellectual faculties, it is not to be supposed that the physical laws to which he is subjected should be essentially different from those which are observed to prevail in other parts of the animated nature.”13 Thus the Malthusian model dominates in evolutionary ecology as well. For animal and plant species population equilibrium is similarly attained when birth rates equal death rates. Birth and death rates are both assumed to be dependant on the quality of the habitat, the analogue of the human level of technology, and population density. Ecological studies typically consider just the direct link between birth and death rates and population density, without considering the intermediate links, such as material consumption, as I have done above. But the Malthusian model for humans could also be constructed in this more reductionist way. At least some ecological studies find that population density affects mortality in ways that are analogous to those we have posited for human populations, through the supply of food available per animal. Thus one study showed that over forty years wildebeest mortality rates depended largely on the available food supply per animal: “the main cause of mortality (75 percent of cases) was undernutrition.”14 Hence the Industrial Revolution after 1800 represented the first break of human society from the constraints of nature, the first break of the human economy from the natural economy. Political Economy in the Malthusian Era Malthus’s essay was written in part as a response to the views of his father, who was a follower of the eighteenth-century Utopian writers William Godwin and the Marquis de Condorcet. Godwin and de Condorcet argued that the misery, unhappiness, and vice so common in the world were the result not of unalterable human nature but of bad government.15 Malthus wanted to establish that poverty was not the product of institutions, and that consequently changes in political institutions could not improve the human lot. As 13. Malthus, 1830, 225. 14. Mduma et al., 1999, 1101. 15. Godwin, 1793; Condorcet, 1795.                               we have seen, in a world of only episodic technological advance, such as England in 1798, his case was compelling. Certainly one implication of the Malthusian model, which helped give classical economics its seemingly harsh cast, was that any move to redistribute income to the poor (who at that time in England were mainly unskilled farm laborers) would result in the long run only in more poor, perhaps employed at even lower wages. As Ricardo noted in 1817, “The clear and direct tendency of the poor laws is in direct opposition to these obvious principles: it is not, as the legislature benevolently intended, to amend the condition of the poor, but to deteriorate the condition of both poor and rich.”16 The poor laws would lower wages because they aided in particular those with children, thus reducing the costs of fertility and driving up the birth rate. The arguments of Malthus and his fellow classical economists not only suggested the inability of government to improve the human lot through traditional methods, they also implied that many of the government policies that the classical economists attacked—taxation, monopolies, trade barriers such as the Corn Laws, wasteful government spending—would similarly have no effect on human welfare in the long run. But the classical economists did not see this. Indeed, if we follow the logic laid out here, good government in the modern sense—stable institutions, well-defined property rights, low inflation rates, low marginal tax rates, free markets, free trade, avoidance of armed conflict—would either make no difference to material living standards in the Malthusian era or indeed lower living standards. To take one example, suppose that the preindustrial king or emperor levied a poll tax on every person in the economy, equivalent to 10 percent of average income. Suppose also that, as was the wont of such sovereigns, the proceeds of the tax were simply frittered away on palaces, cathedrals, mosques, or temples; on armies; or to stock a large harem. Despite the waste, in the long run this action would have no effect on the welfare of the average person. To understand why, refer back to figure 2.1. The tax would act like a shock to the technology of the economy, shifting the lower curve uniformly 16. McCulloch, 1881, 58. Thus classical economics was influential in creating the draconian reforms of poor relief in England in 1834. The most influential member of the Poor Law Commission set up to examine the workings of the old Poor Law was Nassau Senior, professor of political economy at Oxford University.    left by 10 percent. Initially, with the existing stock of people, the tax reduces incomes per person by 10 percent, thus driving up death rates above birth rates. But in the long run after-tax incomes must return to their previous level to stabilize population again. At this point population is sufficiently smaller that everyone earns a high enough wage that, after paying the tax, they have sufficient funds left over to equal their old pretax earnings. In the long run exactions by the state have no effect in the Malthusian economy on welfare or life expectancy. Luxury, waste, extravagance by the sovereign—all had no cost to the average citizen in the long run! Restrictions on trade and obstructive guild rules were similarly costless. Thus at the time the Wealth of Nations was published in 1776, when the Malthusian economy still governed human welfare in England, the calls of Adam Smith for restraint in government taxation and unproductive expenditure were largely pointless. Good government could not make countries rich except in the short run, before population growth restored the equilibrium.17 So far we have considered only actions by government that shift the effective consumption possibilities for a society. Governments could also directly affect birth rates and death rates through their policies. War, banditry, and disorder all increased death rates at given levels of income (though war often killed more through the spread of disease than from direct violence). But all increases in death rates make societies better off in material terms. Here “bad” government actually makes people better off in material terms, though with a reduced life expectancy. Good governments—those that, for example, store grains in public granaries as a hedge against harvest failures, as in some periods in Imperial Rome and late Imperial China—just make life more miserable by reducing the periodic death rate from famines at any given average material living standard.18 It is thus ironic that—while the classical economists, and in particular Adam Smith, are taken as their intellectual fathers by modern proponents of limited government—their views made little sense in the world in which they were composed. 17. It is explained in chapter 5 that high incomes in eighteenth-century England probably owed more to bad personal hygiene than to advances in political economy. 18. In China state granaries in the eighteenth century routinely distributed grain to the poor. See Will and Wong, 1991, 482–83.                               Income Inequality and Living Standards Preindustrial societies differed in their degree of income inequality. Based on modern evidence, forager societies were egalitarian in consumption. In such communities there was no land or capital to own, while in settled agrarian societies as much as half of all income could derive from ownership of assets. Furthermore, forager societies were typically characterized by a social ethic that mandated sharing. Thus, for example, even the labor income of successful hunters was taxed by the less successful. Agrarian societies from the earliest times were much more unequal. The richest members of these societies commanded thousands of times the average income of the average adult male. Aristocrats, such as the Duke of Bedford in England in 1798, resided in a state of luxury that the farm laborers on his extensive estates could hardly comprehend. The Malthusian model takes no account of income distribution. But, by analogy with the discussion of the previous section on taxation and living standards, we can see that greater inequality will have little or no effect on the living standards of the landless workers, the mass of the population. The more equally land rents and capital income are distributed across the general population the more these rents will simply be dissipated in larger population sizes. If these rents were instead appropriated by an aristocratic elite, as they were in many preindustrial societies, then they could be enjoyed with little or no cost to the rest of the population. Thus while inequality could not make the median person better off in the Malthusian world, it could raise average income per person by raising the incomes of the propertied elite. Thus it was possible that England, France, or Italy in 1800 could have a higher income per person than the original foragers. But perversely they would achieve this higher income only through their achievement of greater inequality than earlier societies. And the boost to income per person from inequality was limited. Land rents and capital income made up perhaps half of all income in settled agrarian societies. The expropriation of all these incomes by an elite would double income per person compared to a state of complete inequality. In summary table 2.2 shows Malthusian “virtues” and “vices.” But virtue and vice here are measured with reference only to whether actions raised or lowered material income per person.19 19. Chapter 3 explains why indolence is a virtue in Malthusian economies.    Table . Malthusian “Virtues” and “Vices” “Virtues” “Vices” Fertility limitation Bad sanitation Violence Harvest failures Infanticide Income inequality Selfishness Indolence Fecundity Cleanliness Peace Public granaries Parental solicitude Income equality Charity Hard work The Neolithic Revolution and Living Standards The great economic transformation of the preindustrial era was the Neolithic Revolution: the move from hunter-gatherer societies to those with economies based on cultivated crops and domesticated animals. Anthropologists and archaeologists have long debated what effect this transformation had on living standards, with many believing that farming reduced them. Jared Diamond has even gone so far as to argue that “Forced to choose between limiting population or trying to increase food production, we chose the latter and ended up with starvation, warfare, and tyranny.”20 The empirical data are inconclusive. We shall see in chapters 3–5 that the evidence, on balance, is that living standards in the broadest sense— consumption, leisure, life expectancy—did decline after the spread of settled agriculture, but with significant variation across different agrarian societies. We will find in the following chapters that these modest declines are explained by the fact that the birth rates of forager and settled agrarian societies were likely the same, and death rates at a given income differed little. The ability to store food in settled agrarian societies, which allowed for survival of lean periods and so reduced death rates, would reduce living standards. On the other hand, increased disease mortality from greater population densities helped increase material living standards. The net result of these effects could go either way. Thus the effect of settled agriculture on living standards in a Malthusian world is inherently ambiguous. 20. Diamond, 1987, 66. See also Cohen, 1977; Kaplan, 2000.                               The failure of settled agriculture to improve living conditions, and the possibility that living conditions fell with the arrival of agriculture, have led some economists, anthropologists, and archaeologists to puzzle over why mankind abandoned the superior hunter-gatherer lifestyle for inferior agrarian societies.21 But within the framework of the Malthusian model there is no puzzle. Agriculture was adopted because it was initially a better technology, which generated higher incomes. But those higher incomes inevitably led to larger populations and a decrease of living standards to a new Malthusian equilibrium, seemingly one less favorable than that for the previous huntergatherer societies. Material Conditions: Paleolithic to Jane Austen This chapter explained the first claim made in the introduction, that living standards in 1800, even in England, were likely no higher than for our ancestors of the African savannah. Since preindustrial living standards were determined solely by fertility and mortality, the only way living standards could be higher in 1800 would be if either mortality rates were greater at a given real income or fertility was lower. This conclusion may seem too powerful in the light of figures 1.1 and 1.2. But the upper class about whom authors such as Jane Austen wrote were a small group within English society. In Sense and Sensibility one of her characters says, of an income of £300 a year from a rectory, “This little rectory can do no more than make Mr. Ferrars comfortable as a bachelor; it cannot enable him to marry.”22 In contrast the mass of farm laborers in England in 1810 had an annual income of £36 or less per year. Even though England was one of the richest economies in the world, its people lived by modern standards a pinched and straightened existence. If employed they labored three hundred days a year, with just Sundays and the occasional other day off. The work day in winter was all the daylight hours. Their diet consisted of bread, a little cheese, bacon fat, and weak tea, supplemented for adult males by beer. The diet was low in calories given their heavy manual labor, and they must often have been hungry. The monotony was relieved to some degree by the harvest period, in which work days were long 21. See, for example, Cohen, 1977, and Richerson et al., 2001. 22. Austen, 1957, 247.    but the farmers typically supplied plenty of food. Hot meals were few since fuel for cooking was expensive. The laborers generally slept once it got dark since candles for lighting were again beyond their means. They would hope to get a new set of clothes once a year. Whole families of five or six people would live in two-room cottages, heated by wood or coal fires.23 Almost nothing they consumed—food, clothing, heat, light, or shelter—would have been unfamiliar to the inhabitants of ancient Mesopotamia. Had consumers in 8000 BC had access to more plentiful food, including meat, and more floor space, they could easily have enjoyed a lifestyle that English workers in 1800 would have preferred to their own. In the following three chapters I show that all the major empirical implications of the Malthusian model hold true for the world in the years before 1800. 23. Eden, 1797; Clark, 2001b.                              