The Rise and Decline of Nations

When testing any theory, it is by no means enough to find a few cases that seem to support it. If the predictions of the theory apply to a large number of cases and only a small number of those cases are discussed, then there is always the possibility that only those cases that happen to be consistent with the theory have been considered and that a thorough analysis of the available evidence would indicate that the theory was false. Unfortunately, the present theory has implications for such an incredibly wide array of phenomena in different countries and historical periods that a thorough and meticulous testing is out of the question here. It would not only make this book impossibly long, but it would also require vastly more knowledge than I have or could hope to acquire. There is, however, protection against the possibility that I have considered only unrepresentative cases in the fact that we have looked at the growth rates of all of the developed democracies in the years since regular estimates of national income were first prepared, shortly after World War II (see table 1.1, p. 6 above). Thus, so far as the postwar developed democracies are concerned, there is no possibility that only those cases that happen to fit the theory have been considered. It may be a matter of dispute whether the theory is consistent with all of the major variations in growth rates in this subset or just with most of them, but since the importance of other causal factors is emphasized, there is some support for the theory in the postwar experience of the developed democracies on even the most skeptical reading of this evidence.

It is, on the other hand, still possible that the developed democracies since the war are unrepresentative in ways that are crucial to deciding what claim to credence, if any, the theory has. One real possibility is that the data on these countries fit the argument purely by chance, although the collateral evidence on industry comparisons and temporal patterns of growth and the mass of corroborative data on the forty-eight states make this extraordinarily unlikely. Another possibility is that there is some altogether different causal mechanism operating that has much the same results as the theory here predicts. A third possibility is that the theory is true or largely true for these countries but does not apply to other types of societies, such as the developing nations or the communist countries.

There is some protection against all these possibilities in the way the theory appears to fit the experience of Britain, Holland, and France in the early modern period, the patterns of growth within European countries in that period, and the growth of the United States and Germany in the nineteenth century. Indeed, as we shall see in the last chapter, the theory is consistent with some dramatic features of the interwar period also, so the claim that it fits the most striking departures from normal economic experience among the nations of the West since the late Middle Ages has some basis.

No matter how much additional evidence might substantiate the findings on the developed democracies and on the modern economic history of the West, it still would not support compelling conclusions about non-Western societies (except perhaps Japan). Possibly different causal processes are operating in non-Western areas. Erudite scholars like Max Weber have argued that certain features of Western European Christendom, and especially of puritanism, were uniquely favorable to capitalism and economic progress. Although the historical support for Weber’s fascinating argument is at best mixed, we must ask whether the coalitional processes described in this book are dependent on certain cultural or religious attitudes and confined more or less to Western civilization. We should not conclude that the same tendencies are at work in other civilizations unless there is evidence of similar coalitional processes in several other cultural traditions.

There is, in fact, massive evidence of coalitional processes in a variety of non-Western societies. There have been guilds, for example, in Moslem countries (and even in Mecca),¹ in Byzantium, in China, in Hellenistic times, and even in Babylonia.² These guilds, moreover, bear the same dead-giveaway signs of cartelistic purposes: restrictive membership, price-fixing, long apprenticeships from which the sons or relatives of members are often exempt, and rules limiting output and innovation. As might be expected from the many modem studies finding similar motivations and responses to prices and profit opportunities in developed and developing nations,³ the eager acceptance of the gains from cartelization and political power seems much the same in very different cultural and religious traditions. Whereas those parts of the world that have never developed very far cannot show us quite the contrasts that European economic history offers, it is still clear that guilds and other distributional coalitions have normally had the same harmful effects on economic efficiency and growth, whatever the culture.

The Chinese economy in the latter part of the nineteenth century offers a good example. Even though some guilds had been destroyed in the instability associated with the Taiping rebellion in the mid-nineteenth century, guilds were powerful, especially in the latter part of the century. Hosea Ballou Morse, a leading scholar on China (and Commissioner of Chinese Maritime Customs for a time during the *'treaty ports” period), wrote in The Guilds of China (1909) that “all Chinese trade guilds are alike in interfering with every detail of business and demanding complete solidarity of interest of their members, and they are alike also in that their rules are not a dead letter, but are actually enforced. The result is tyranny of the many over the individual, and a system of control which must by its nature hinder freedom of enterprise and independence of individual initiative.”⁴

Some economists argue that there cannot be much monopoly or cartelistic power unless the coercive power of government is brought into play, but Chinese guilds provide unusually clear evidence that this view is wrong. There were, to be sure, symbiotic relations between guilds and government officials in which the coercive power of government was brought to bear in the common interests of the guild and the officials. Nonetheless, as Morse stated it, “The trade guilds have grown up apart from and independent of the government; they … devised their own regulations, and enforced them in their own way and by their own methods.”⁵ He argued that Chinese guilds could enforce their regulations.

Partly because of the enormous impulsive power of a mediaeval form of public opinion and the development of the boycott by centuries of practical use, the guilds have in fact obtained an enormous and almost unrestrained control over their respective trades…. Their jurisdiction over their members is absolute, not by reason of any charter or delegated power, but by virtue of the faculty of combination by the community and of coercion on the individual…. The craftsman who is not a guild member is as one exposed to the wintry blast without a cloak.⁶

Even individual appeals to the government on matters of interest to the guild were excluded, unless the guild had first considered the matter. Another observer, Daniel J. Macgowan, cites guild rules specifying that if a ‘'complainant have recourse to the [government] official direct, without first referring to the guild, he shall be subject to a public reprimand, and any future case he may present for the opinion of the guild will be dismissed without a hearing.”⁷

Guild power could be used even against the government. There is a ghastly illustration of this in Macgowan’s reports of the gold beaters guild, which provided gold leaf that the emperor purchased in large quantities. The rule of the trade was that no employer could have more than one apprentice at a time, but one member of the craft represented to the magistrate that, if he were allowed to take on a number of apprentices, the work would be expedited. He received permission to do so and engaged a great many apprentices. This output-increasing, price-reducing conduct infuriated the craft. The word was passed around that ⁴‘biting to death is not a capital offence,” apparently on the gruesome theory that no one of the morsels taken is fatal, and the cartel-buster was soon dead from the fiendish efforts of 123 of his fellows.⁸ None of the guild members was allowed to leave the shop until his teeth and gums attested to what, in more delicate settings, might be called his ‘'professional ethics.” Although the man who took the first bite was, it turns out, discovered and executed, one can well imagine that the squeamish, at least, must have been made apprehensive about increasing output or cutting prices, even when the emperor was the buyer.

The effect of guilds was no doubt increased by the fact that China, ostensibly a unified nation, had tariffs or transit taxes on trade within the country.⁹ In addition, there was effectively a prohibition against foreign trade (because all imports had to go through a single guild in Canton), until the Western powers imposed treaties on the Chinese that opened certain ports and commercial opportunities to foreigners. After these treaties there were many efforts to introduce various types of production with modern Western technology, many of which were defeated by boycotts or governmental discouragements organized by guilds of competitors. Guilds blocked or delayed the use of modern technologies in silk reeling, coal mining, soybean-oil pressing, steamboat transportation, and railways, for example. Chinese as well as foreign businessmen were discouraged from investing in new technologies, and the most successful Chinese businessmen were concentrated in the treaty port cities under European jurisdiction.¹⁰

China, though possessed of an extraordinarily ancient and rich culture, did not of course industrialize. Only a few decades ago it was often taken for granted, even by the most erudite and sympathetic observers, that something in the Chinese spirit or culture was inherently unsuited to modern economic life.¹¹ This is nearly the opposite of the conventional wisdom now. In the last three decades the most rapidly growing areas in the world have been Chinese or profoundly influenced by Chinese culture. Consider the communities that I. M. D. Little has called the “gang of four”: Hong Kong, Korea, Taiwan, and Singapore. All four, it is worth noting, have recent histories that have been inimical to the development of distributional coalitions and have had relatively liberal trade policies as well. Korea and Taiwan did not have the freedom to develop independent interest groups while they were colonies of Japan, Singapore had little to gain from lobbies when it was run by Britain, and Hong Kong is still a colony run along nineteenth-century British free-trade lines.¹²

III

Western observers usually greatly underestimate the differences between Chinese and Japanese cultures, but Japan can nonetheless also be considered an area that has been profoundly influenced by Chinese culture. The rapid growth of Japan since World War II has already been analyzed, but there is also the exceptional growth Japan enjoyed for a couple of generations after the Meiji restoration in 1867-68. This earlier phase of Japanese growth also stands in sharp contrast to what occurred in China, and for that matter to what happened at that time in all other non-Western areas of the world.

Many accounts of Japanese growth attribute it mainly to special characteristics of the Japanese culture or people. The Japanese were not, however, always considered economic supermen. Western visitors in the mid-nineteenth century were often struck with the utter poverty of the people and even with the number of families that were reduced to infanticide. Although the rate of literacy was quite high by the standards of poor societies at that time, and the society had been progressing in certain respects,¹³ it was pathetically weak both technologically and militarily and subject to humiliation by even the most casual efforts of Western navies. In those days the conventional wisdom among Western observers was far different than it is today, with some alleging that Japanese character or culture was intrinsically incapable of economic development.¹⁴

Before Admiral Perry’s gunboats appeared in 1854, the Japanese were virtually closed off from the international economy; foreign trade was largely confined to one port and trade through that port was severely limited. A central government of sorts under the shogun had maintained peace and stability for several centuries, but much of the power to determine economic policies remained in the hands of more than 200 separate daimyo, or feudal lords. The tolls, tariffs, regulations, and legal monopolies of these separate fiefs, with their own coinages and currencies, drastically limited trade within Japan.

As we know, the theory offered here predicts that protected markets enjoying a period of stability will become cartelized, at least if the number of enterprises in the market is small enough to allow each individual enterprise to get a significant share of the gain from collective action. This prediction fits Japan no less than other societies; there were any number of powerful za, or guilds, and the shogunate or the daimyo often strengthened them by selling them monopoly rights. Various guilds controlled major markets, although there were also independent enterprises in rural areas and even merchants who used the ‘ ‘putting-out” system. Of course, Japanese guilds fixed prices, restricted production, and controlled entry in essentially the same way as cartelistic organizations elsewhere.

The reader may be weary of seeing the same story over and over again in different settings, but since the causes of Japanese growth are shrouded by tenacious myths, it is perhaps best to be explicit. The upheaval that led to the Meiji restoration not only deposed the shogun and effectively dispossessed many of the vested interests tied to the shogunate, but soon also abolished the domains of the feudal daimyo as well and all of the restrictions on trade and enterprise that went with them. At about the same time that the Meiji government eliminated the barriers to a national market, Britain and other Western powers imposed treaties upon the Japanese that required something approaching free trade with the rest of the world. In particular, a treaty of 1866 restricted the Japanese to a revenue tariff of not more than 5 percent, which lasted until 1899. It was the military, technological, and economic weakness of the Japanese that forced them to accept the provisions of this and similar agreements, which the Japanese are accustomed to describing as “humiliating.”

Lo and behold, the Japanese were humiliated all the way to the bank. Trade immediately expanded and economic growth apparently picked up speed, particularly in the 1880s and 1890s, and just after the turn of the century a new Japan was able to triumph in the Russo-Japanese War. Once again, multiple causation must be emphasized. For example, the government subsidized industries that were deemed important for military purposes and also promoted education effectively. Quantitatively speaking, however, the overwhelmingly important source of Japanese growth in the nineteenth century was the progress of small-scale private industry and agriculture, such as exports of silk and tea. Interestingly, most of the important Japanese entrepreneurs in this period do not trace their origins to the merchant houses belonging to the guilds of the pre-Meiji period; but rather, they came disproportionately from the ranks of impoverished lesser samurai (who, by the precepts of traditional Japanese culture, were not supposed to engage in commerce at all) or from rising farm and trading families in rural areas that were more likely to be beyond control by guilds or officials. It is said that when markets opened up, many of the houses that had belonged to guilds were disoriented and at a loss what to do.¹⁵

It is natural to turn from East Asia and the countries that have been most influenced by China to South Asia and particularly to India. Like China, India has an unusually ancient history, a rich culture, and a huge impoverished population. Yet there are also colossal and oft-neglected differences between these two countries. China was among the earliest, if not the earliest, of the nation-states, and, in spite of the several occasions when its empires have collapsed, it has an extremely long history as (more or less) a single country. India, of course, did not have a single government over the whole of the subcontinent, or over all of what is now India, until it fell under British control. The Indian subcontinent is also geographically divided, by deserts, jungles, and mountains, to a greater extent than the populous parts of China are. The conquest of formative areas of Indian civilization by Aryan-speaking peoples about 1500 B.C. also introduced a further disparity into Indian life that has no counterpart in Chinese history—the Mongol conquerors of China did not impose their religion, for example, on Chinese society but were instead thoroughly assimilated by it. For these and no doubt other reasons, India is in several important respects more diverse than China. It does not have a single language common to its many peoples, whereas China (despite the vast differences in its dialects) has at least a common written language. Even a glance at the physical appearances of people from the two nations suggests much less diversity among the Chinese than among the Indians. This great diversity suggests that it is wise to be skeptical about any generalizations concerning India, including those that will be offered here.

The mosaic of jurisdictions that covered the Indian subcontinent in the pre-British era changed many times. Thus there was often a good deal of instability and war, as some warlords or dynasties expanded and others retreated or were defeated. In many periods of Indian history, however, there was what the British called “indirect rule.” The British in India and throughout their empire usually did not seek to impose their government down to a local level, much less require every community or tribe to follow uniform rules. They would often rule indirectly by letting traditional authorities, decision-making arrangements, and customs prevail, provided there was no insurrection or outrage against British sensibilities or interests. As time went on or conditions changed, there might be somewhat more obtrusive government, but the British never attempted to eliminate the traditional religion or social structure of India or to remake all of Indian society along British lines; they deliberately kept out all missionaries, for example, until 1813.¹⁶ Indirect rule was also characteristic of the Moghuls who earlier ruled the more northerly parts of India, although some of these rulers did encourage or require conversions to Islam. The Moghuls did not have bureaucracies akin to those of modern governments, or even to those of the Chinese emperors, and could not impose detailed or uniform government at a village level. Often supporters were given dijagir (a right to tax a collection of villages), but they would not normally own land or manage the daily life in these villages. Sometimes Hindu nobles, or zamindars, retained hereditary control of village revenues, and some Hindu princes continued to rule and collect taxes in autonomous states within the Moghul empire.¹⁷ The diverse rulers of the various parts of India before the Moslems did not appear to have the bureaucracy and efficiency needed to administer vast areas in a uniform way and also appear to have taxed villages as units rather than separately taxing the individuals in the village. In general, traditional India did not have individual ownership of land, and both the different groups in the village and the rulers would share in the output resulting from the cooperative effort and division of labor in the village. At local and especially village levels, then, life could often continue without great change or instability even when new rulers came to exact taxes and tribute.

My thoughts about how the theory offered in this book relates to India occurred when, quite by chance, I was reading Jawaharlal Nehru’s The Discovery of India. This remarkable book was written in only five months in 1944, while Nehru was confined by the British to Ahmad-nagar Fort prison. Even though I had previously read accounts of how profoundly Nehru had been influenced by his English education and experience, I nonetheless expected that Nehru—who was after all a political figure who already may have hoped to become the leader of an independent India—would celebrate the glories of India’s ancient civilization and place almost all of the blame for the country’s problems on the British. Nehru naturally did point with pride to many of the great achievements of India and Indians, but what was most notable was that his praise was confined almost exclusively to certain periods of Indian history, whereas Indian society and institutions in other periods were criticized if anything more seriously than he criticized the rule of his British jailers.

Nehru was impressed, as everyone must be, by the precocious civilization in the Indus valley, one of the world’s earliest societies with settled agriculture and what can fairly be called civilization. He cites, for example, the impressive houses, baths, and drainage systems evident in the excavations of the ancient city of Mohenjo-daro and quotes Western authorities who compare aspects of this Indus valley civilization favorably with contemporary civilizations in Egypt and Meso potamia. He also points out that what the West calls “Arabic” numerals came originally from early India and that the discoveries of the concept of zero in a number system and of symbolic, algebraic notation were further examples of the creativity of early Indian civilization. Western Europe in the Dark Ages, he argues, was backward by the standards of Asia at that time.

He offered a very different view of Indian civilization at the coming of the Moslems and at the conquest of India by the Europeans. India in this epoch was “drying up and losing her creative genius and vitality;” it was the “afternoon of a civilization.” This “stagnation and decay” was pervasive: “There was decline all along the line—intellectual, philosophical, political, in technique and methods of warfare, in knowledge of and contacts with the outside world, in shrinking economy.” It was true, Nehru conceded, “that the loss of political freedom leads inevitably to cultural decay. But why should political freedom be lost unless some kind of decay has preceded it? A small country might easily be overwhelmed by superior power, but a huge, well-developed and highly civilized country like India cannot succumb to external attack unless there is internal decay, or the invader possesses a higher technique of warfare. That internal decay is clearly evident in India.” Most of the above quotations from The Discovery of India relate to India at the time much of it was conquered by the Moslems, but he is clear that the same stagnation was evident when the Europeans conquered India a few centuries later and points out that they were able to capture the subcontinent “with remarkably little effort;” there was “a certain inevitability in what happened.”

Nehru attributed the decay to “the static nature of Indian society which refused to change in a changing world, for every civilization which resists change declines.” He reasoned that “probably this was the inevitable result of the growing rigidity and exclusiveness of the Indian social system as represented chiefly by the caste system.” The caste system, he wrote, was a “petrification of classes” that “brought degradation” and is “still a burden and a curse.”¹⁸ Nehru did not claim any originality for this diagnosis and it is fairly common. There are also limits to the reliance that can be placed upon the hurried writings of a jailed political leader without the best access to sources and specialists. Nonetheless, I think that the account must have resonated with the experience and observations of many educated Indians, for the book as well as the author have been widely celebrated in India as elsewhere.

Quite apart from its effects on efficiency, the caste system is also a source of profound inequality, both in opportunities and in results. In India today there have been changes, but one must remember that in the traditional caste system groups in the population were condemned for life, and their descendants in perpetuity after them, to such tasks as the cleaning of latrines and the removal of dead carcasses. Their very touch, and in some cases even their nearness, was deemed to be polluting, their presence in temples defiling. Apart from slavery, it is hard to think of a system with greater inequality of opportunity, and the results are also most unequal. This inequality was, of course, also of great concern to Nehru.

It is not sufficient to explain the decline of India by the era of the Moslem and European invasions in terms of the caste system. That, too, is an untestable ad hoc explanation; Indian history is unique in countless ways, and there is no way to determine whether a given unique trait was in fact the source of the decline Nehru noted. We have not reached home until we have explained why India acquired the caste system when it did and have comprehended the caste system in a theory that is testable against the experience of many countries.

The sources for the distant past of India are scanty and so little is known that agnosticism is very much in order. There is general agreement that India did not always have the caste system. It is not normally thought to have been part of the civilization in the Indus valley that preceded the Aryan conquests. Neither do the Vedas of the Aryan invaders speak of the ritual purity and pollution or the prohibitions against intermarriage or change in rank that characterized the caste system. How then did the caste system emerge?

One of the most common hypotheses is that the castes emerged out of guilds or similar organizations; most castes bear the names of occupations and there is evidence of guilds in earlier Indian history. Another common hypothesis is that visible racial differences among the indigenous peoples of India and between these peoples and the Aryan-speaking invaders were the source of the caste system; there are visible differences among some caste groups to this day, and the English word caste stems from the Portuguese casta, meaning race. Yet another familiar explanation ascribes the castes to common descent; a crucial feature of the caste system is endogamy or the prohibition against marriage outside the basic unit of caste grouping, the jati, and many tribes have been incorporated into the caste system. It might seem that the theory here would focus exclusively on the hypothesis that the castes grew out of guilds, but the other two hypotheses are also important, if the theory offered here is right, and we shall return to them shortly.

Castes traditionally have behaved like guilds and other distributional coalitions. With modernization many new occupations have emerged and the caste system has changed for other reasons as well, so the caste need not be primarily an occupational or guild-type classification for the educated Indian today. Traditionally, however, caste groups were not only mainly occupational, but also exhibited all the features of cartels and other special-interest organizations. They controlled entry into occupations and lines of business, kept craft mysteries or secrets, set prices monopolistically, used boycotts and strikes, and often bargained on a group rather than an individual basis.

The caste system also had several features that would be expected of distributional coalitions. One of these is that often groups rather than individuals change status. A caste group that enjoys prosperity will rise gradually to a higher status and also may decide collectively to adopt more restrictive ritualistic rules, thereby rising even in terms of the religious concepts of purity and pollution. Another feature is that Hinduism emphasizes the concept of dharma, the duties appropriate to the caste or group. Morality, in other words, is defined not in a universalistic way, but in terms of obedience to the rules of one’s caste or station, so it is similar to professional ethics that rule out competition in a profession. Even the murderous thugs or other criminal castes were behaving consistently with their dharma when they carried on their caste’s activities. A reward for fidelity to the rules of the caste or group into which one is born is a favorable reincarnation. Finally, the one way in which those born into humbler castes can rise in religious status during one lifetime is by leaving the system of group competition and forgoing material satisfactions and affiliations; higher religious status, such as that of Brahmins, is associated with privilege, and any rise in religious status that does not involve renunciation threatens other groups.

VII

None of the preceding, however, is an explanation of the prohibition against marriage out of the group that is such a basic feature of the caste system, nor does it explain any correlation of caste with racial or ethnic differences. For that we must turn to Implication 8, which is that distributional coalitions are characteristically exclusive and seek to limit the diversity of their memberships. We must ask how that implication would apply over a multigenerational time span.

Consider the situation of an older member of a profitable guild. As one of the co-owners of an advantageous coalition, the older member would have an interest in how he or his descendants might share in the future returns. One logical possibility is that he could upon his death or retirement bequeath his share of the future returns of the coalition to his children; his son, for example, could take his place in the craft. But some of the members of the coalition will have daughters and some only daughters. Suppose that the coalition is a guild with only male workers, and that the members with daughters then offer access to the profitable cartel as part of a marriage bargain with sons-in-law. That will offer the old member a way of getting something for his share of his coalition’s worth, but we must ask what will happen if both sons and sons-in-law enter the trade. Even with a steady-state population the number in the craft will double if both sons and sons-in-law are allowed to enter, and normally a doubling of the craft’s membership would eliminate the gains of the cartelistic output restriction that gave the guild its value in the first place. The same problem will occur if both sons and daughters practice a craft that was previously restricted to one sex. The multi-generational guild can be successful only if it can keep its membership from increasing faster than can be justified by any expansion in its market, which will depend on such things as the growth of population and income in the areas in which it is located. Unless some sons are left out, the only way those members who have only or mainly daughters can gain their share of the value of the cartel, without making the cartel valueless, is to restrict the sons allowed to enter the trade to marriages with daughters of members of the trade.

The same is true of a coalition that has, say, disproportionate rights to the village harvest. The greatest distributional gains come from a minimum winning coalition. Thus if the favorable share of the harvest is divided up among more families, there is less for each family. But if each family has on average two surviving children who marry, then there will be two families in the next generation for every family in the first generation, and in a few generations even the grandest entitlement will provide very little per family. The only way the distributional coalition can retain its value over several generations is by restricting the children of members to marriages with one another or by disinheriting a large portion of the children. I hypothesize that the Indian castes mainly used the first method. The English nobility used this method to a great degree and combined it with primogeniture as well (thus it is not astonishing that some great fortunes were passed on for several generations in the English aristocracy).

This reference to the British nobility brings us back to the discussion of Implication 8 in chapter 3, where the nobility and royalty of Europe were used to illustrate the exclusiveness of marriages (or bequests) that is essential to any successful multigenerational special-interest group. Those who have a chauvinistic turn of mind or who are convinced that fundamentally different processes must operate at different levels of wealth or status no doubt will be surprised by this alleged similarity of motivation in groups as different in wealth and background as the European nobility and the Indian castes. Those who have done a lot of empirical and historical research in economics would, I think, be surprised by anything else. Those who have studied barriers to in-tergenerational mobility across social classes in any societies with significant class barriers will also, I conjecture, find incipient castes.

Just as the origin of the caste system is often ascribed to guilds, so is it often related to the racial diversity of India at the time of the Aryan migrations and also to descent groups. Given the repeated emphasis on multicausality and the complexity of reality in this book, we should examine these hypotheses sympathetically. They are, as it turns out, also very much in keeping with the logic behind Implication 8. If a racially distinct distributive coalition is formed by alien conquerors, it will be able to preserve itself over many generations only by arbitrary rules of bequest such as primogeniture or through endogamy. If it is largely endogamous the differences in appearance will be preserved.

Indeed, it will be far easier for a racially, linguistically, and culturally distinctive group to maintain a multigenerational coalition. The linguistic and cultural similarities will reduce differences in values and facilitate social interaction, and, as chapters 2 and 3 show, this reduces conflict and makes it easier to generate social selective incentives. Moreover, any special-interest group that uses endogamy to preserve its benefits over a multigenerational period must be large enough to avoid inbreeding. As the endogampus group gets larger, however, the difficulty of enforcing endogamy rises. How is this or that son to be restrained from marrying some especially appealing girl outside the group, or how are his parents to be prevented from making an especially advantageous marriage contract for him with relatively wealthy or powerful people outside the group? How can the astute outsider be kept from marrying into the coalition? If exogamy is not prevented, at least some of the families must lose their share of the coalition’s future gains. If there are visible differences, it will be easier to determine who is in the group and who is not and to enforce the endogamy rule. Differences in speech, culture, and lifestyle are also shibboleths that make it harder for the outsider to blend in. Unfortunately, the promotion of prejudices about race, ethnicity, culture, and intergroup differences in lifestyle will also make the coalition work better. The inculcation of these prejudices will increase the probability that the members will follow the rule of endogamy and strengthen selective incentives by interacting socially only with their own group, of their own accord.

Though multigenerational distributional coalitions foster inefficiency, inequality, and group prejudice, it is nonetheless important to realize that some individuals and groups outside the society containing these coalitions may improve their positions by joining that society, even if they enter at the bottom. Tribes without settled agriculture, for example, might in some circumstances have found that they would be better off joining Indian society than by staying out of it, even though they were accorded the lowest status and were victims of special-interest groups to boot. There have been many observations of such assimilation of tribal groups into India’s caste system, and they must help to account for its great diversity.

This diversity, once again, reminds us of the complexity of the matter. Because of this complexity and the limited sources on the early years of the caste system, we must not jump to any conclusions. The hypotheses that emerge from the theory here should be considered primarily a stimulus to further research. There has been no theoretical consensus on caste and class: a fresh perspective could provide some help. It is, I submit, worth doing serious research on whether multi-generational processes of the sort the theory suggests have in fact emerged over the millennia of Indian history. This exceptionally long history of settled agriculture and civilized life was combined for the most part with indirect or parochial rulers who could not or did not challenge the power or usurp all the gains of the distributional coalitions. It was combined also with racial diversity and geographical segmentation of markets favorable to the coalition formation by small groups. If the processes of the kind described above did not occur, then how do we explain what happened?

VIII

In keeping with the scientific principle that the theory that explains the most with the least is most likely to be true, any alternative explanation of the Indian caste system should also be capable of explaining some developments outside India, as the present theory does, or at least parsimoniously explain a good deal more about the caste system. By the same token, the explanation offered here of the Indian caste system will be stronger if the theory in the book explains not only diverse developments outside India, but also developments outside India that are similar to the Indian developments. To some extent the previous analysis of class rigidities was in that category, but there is still the problem that the Indian caste system is unique, so that developments in other countries do not provide the close parallelism we seek. In particular, in most other countries any class rigidities usually do not involve rigid requirements of endogamous marriage and group or race prejudice. Since the theory here implies that over a sufficiently long run distributional coalitions will stimulate group prejudice and promote endogamous marriage, it is wise to look for countries other than India where this is occurring. Other societies will not have the extraordinary antiquity and cultural richness of India, but perhaps some of them will in certain respects resemble India in the era when the caste system emerged. We should particularly look for societies with racial and cultural differences.

The extraordinary system of apartheid in South Africa is a relatively recent development. The more severe forms of racial segregation and discrimination do not go back to the early days of the Boers in South Africa. On the contrary, there was more than a small amount of interbreeding between the Boers or other Europeans and the Africans. There is, after all, a large population of “Coloured” or mixed-race people in South Africa today; the South African government treats them as a separate category and segregates them from Africans and Asians as well as from Europeans.

A distinguished South African economist, W. H. Hutt, in The Economics of the Colour Bar,¹⁹ has written a startling history of the evolution of progressively tighter systems of racial segregation and discrimination in South Africa. Although Hutt is perhaps insufficiently detached about his classical liberal ideology and may offend some readers of other persuasions, my checks with other specialists on South Africa suggest that even those who do not share his interpretation are generally in agreement on his rendition of the historical facts.

Hurt’s account focuses closely on the mining industry in South Africa early in the present century. The mine owners and management needed labor and naturally preferred to secure it at low wages rather than high wages. Since Africans had few other opportunities outside the traditional sector of African society, they were often available at low wages. The mine owners also drew upon the huge pool of low-wage labor in Asia and for a time used indentured Chinese labor. European workers were employed in the mines mainly as foremen and skilled and semi-skilled laborers. It was clear that that the far-cheaper African laborers could at very little cost soon be taught the semi-skilled jobs and the employers naturally coveted the savings in labor cost that this would bring.

The competition of cheaper African and Asian labor did not appeal to the higher-paid workers of European stock or their recently formed unions. There were strikes. In part because of these strikes there were changes in labor policy in South Africa. The Mines and Works Act of 1911, also called the “Colour Bar Act,” was passed. On a superficial reading relatively innocuous, as administered it constrained employers in their use of African labor in semi-skilled and skilled jobs. The regulations promulgated under the act prevented Africans in the Transvaal and the Orange Free State from entering a wide variety of mining occupations. They even specified ratios between foremen (whites) and mining laborers (Africans).²⁰

Disagreement about the ratios emerged. After World War I, the mine employers asked for a ratio of 10.5 Africans per white worker, whereas the labor union demanded 3.5 to 1. A general strike in the Rand followed in 1922. This strike and the agitation that followed became a common cause of conservative Afrikaaners and communist and socialist leaders, with all of them supporting the efforts to deny opportunities to the poorer Africans who were competing with white labor. The South African Labour Party, modeled more or less after its British counterpart, prospered in the wake of the strike and joined with the mainly Afrikaaner, white supremacist Nationalist Party in a coalition government. The Nationalist-Labour “Pact” government soon introduced the second ‘'Colour Bar Act,” the Mines and Works Act of 1926. Hutt calls this ‘ ‘probably … the most drastic piece of colour bar legislation which the world has ever experienced.”²¹ It was accompanied by a “civilized labour policy,” which limited opportunities for Africans still further. One of the devices used to keep African laborers out of jobs where they would compete with whites was the requirement of “the rate for the job.” If the wage for a given job is fixed at a level attractive to Europeans, the employer has no incentive to seek African workers who would work for less. Apprenticeship rules under the “civilized labour policy” also had the effect of excluding Africans.

These and similar policies drastically limited opportunities for African workers. The denial of various skilled and semi-skilled jobs to Africans not only raised the wages of the European (and sometimes Coloured and Asian) workers, but it also crowded more labor into the areas that remained open to Africans, making the wages there lower than they would otherwise be. It is important to remember, though, that there was a continuing demand of Africans from farther north to enter, notwithstanding the policies against them. They came in at the bottom and were victimized by the rules, but it was still better than the alternatives some of them had in the traditional sector. The analogy with the tribes that have been assimilated into the bottom of the Indian caste system is striking.

Since firms that could hire unusually inexpensive African labor had an advantage over foreign or domestic competitors without such opportunities, they would often be profitable even if they were forced to pay more for certain skills because only whites could be employed, to hire more foremen than needed, and so on. There were efforts of firms to move to areas where restrictions on the use of African workers were fewer, but this too was curtailed, as were some African entrepreneurs. Thus the system, while it forced employers to adopt less profitable and more discriminatory policies than they preferred, brought substantial gains to organized white (and sometimes Coloured and Asian) workers.

The theory offered in this book suggests that the employers would have been just as interested in excluding competitors as the workers were, and would as small groups have been better able to organize to do so than the workers. But the competitors of the employers were other firms or capitalists, often in other lands; the employers were not competing against African laborers, as the white workers were, so the employers were not a principal source of the racial exclusion and discrimination that the Africans suffered. South African consumers of all races paid higher prices because of the higher costs growing out of the discriminatory policy, but, as in other countries, they were not organized.

Let us now ask what necessary conditions must be met if the South African system, and the cartelistic gains it provides for many, are to be preserved over the long run. There is a need for police and military power, but this is widely understood and discussed, so it need not be considered here. The system could not possibly survive for many generations unless the demarcation between the races was preserved. If less-favored groups could enter the more-favored groups, as they would have massive incentives to do, wage differentials could not be maintained. A continuation of the processes that generated the Coloured population would make the system untenable in the long run, and even in what (by the standards of Indian history) would be the medium run.

That is not only an implication of the present theory but evidently the conclusion of the South African government as well. Just as the restrictions on the use of African labor in skilled and semi-skilled jobs increased over time, so did the rules separating the population into rigid racial categories and forbidding sexual relations, in marriage or otherwise, between them.

Undoubtedly any number of other causal factors have been at work in South Africa, and any account as brief and monocausal as this must be in many respects misleading. The purpose, however, is not to give a complete account, but to induce reflection on the sources of racial and other forms of discrimination. As others have argued before, the individual as a consumer, employer, or worker finds it costly to discriminate. The consumer who discriminates against stores owned by groups he finds offensive has to pay higher prices or suffer a lesser selection by shopping elsewhere. The employer who discriminates against workers of a despised group has higher labor costs, and his business may even bankrupt itself competing against other firms that do not let prejudice stand in the way of profit. Similarly, the worker who does not accept the best job irrespective of the group affiliation of the employer essentially is taking a cut in pay. A similar logic applies to individual social interactions of other kinds. The fact that individuals find discrimination costly means that, if individuals are free to undertake whatever transactions they prefer, there will be a constraint on the extent of discrimination.

Distributional coalitions of individuals, on the other hand, can sometimes gain enormously from discrimination. Any group difference that facilitates exclusion, by Implication 8, will be advantageous. For periods of only a generation or two in length, the group differences can usually be considered as given, but over the centuries and certainly the millennia they cannot. In the long run, then, multigenerational special-interest groups must tend toward endogamy. This is equally true of the South African whites, the Indian castes, and the European nobility.

This book has not even touched upon societies of the soviet type. The declines in the growth rates of these societies over the stable postwar years are quite as notable as in other countries. Unfortunately, the way the present theory applies to societies of this type cannot be set out briefly; the theory of collective action by small groups needs to be elaborated and the limited role of markets in these societies analyzed. It would be digressive to go into these issues now, and so an account of how the present theory applies to these societies must be left for another publication.

The other class of contemporary societies that has so far been ignored is the characteristically unstable countries. Instability in France and on the Continent were discussed earlier, but nothing has been said about the depressingly large number of less developed countries, in Latin America, Africa, and elsewhere, that have been persistently unstable.

The dense network of distributional coalitions that eventually emerges in stable societies is harmful to economic efficiency and growth, but so is instability. There is no inconsistency in this; just as special-interest groups lead to misallocations of resources and divert attention from production to distributional struggle, so instability diverts resources that would otherwise have gone into productive long-term investments into forms of wealth that are more easily protected, or even into capital flights to more stable environments. On the whole, stable countries are more prosperous than unstable ones and this is no surprise. But, other things being equal, the most rapid growth will occur in societies that have lately experienced upheaval but are expected nonetheless to be stable for the foreseeable future.

The characteristically unstable countries are usually governed part of the time by dictators or juntas; they have intervals of democratic or at least relatively pluralistic government. The policies of the dictators or the juntas obviously will depend dramatically on the interests, the ideology, and sometimes even the whims of the dictator or the leadership group. Experience and common sense tell us that dictators and juntas may be right-wing or left-wing, this or that, although they are systematically more likely to be specialists in violence than in economics. The theory here cannot tell us what policies the dictators and juntas will have. As the better historians remind us, much of what happens in history is due to chance and must remain beyond the explanatory powers of any theory.

Fortunately, something of a systematic or theoretical nature can be said about the influences and pressures that will be brought to bear on the changing governments of the unstable societies, and about their intervals of democracy or pluralism. Implication 3 states that small groups are more likely to be organized than large ones, but that (since small groups organize less slowly) the disproportionate organizational and collusive power of small groups will be greatest in lately unstable societies. The theory here predicts that the unstable society will have fewer and weaker mass organizations than stable societies, but that small groups that can collude more readily will often be able to further their common interests. The groups may be at any level, but usually those which can gain from either lobbying or cartelizing at a national level are small groups of substantial firms or wealthy and powerful individuals.

The tendency for small groups to be better organized than large ones is further accentuated in unstable societies by two other factors. One is that large groups are more likely to be a threat to a dictator or a junta than small ones. If there were an association that included most of the peasants, or a labor union that represented most of the workers, in an unstable country with a dictator, that organization could pose a threat to the dictator. The sheer numbers of the membership of the mass organization would give it some coercive power. (Actually, this point holds for undemocratic regimes whether they are unstable or not—even the stable totalitarian state does not like the threat inherent in independent mass organizations. Unobtrusive small groups accordingly also play a leading role in the application of the theory to totalitarian societies.)

The second factor is that the small group can be discreet and inconspicuous during the dictatorial periods, whereas the organized large group cannot. Dictators, juntas, and totalitarian leaders are not enthusiastic about independent organizations of any kind, but they cannot repress collusions they do not know about. The likelihood that a group will be exposed by an indiscretion of one of its members must rise with its size and becomes a virtual certainty with a large group. Thus if a small group should feel threatened in a repressive society, it can often retain its coherence by becoming invisible to the authorities and then be able to act promptly when there is relaxation or elimination of the repression. (This point, too, applies to stable despotisms as well as unstable ones.)

The most basic implication of the theory for unstable societies, then, is that their governments are systematically influenced by the interests, pleas, and pressures of the small groups that are capable of organizing fairly quickly. Admittedly, the policies of unstable countries may shift wildly, with each coup d’etat bringing new policy preferences. The economic policy of such countries is similar to a leaf blowing in the wind—a gust may blow it suddenly in any direction, but over time gravity still will pull it to the ground.

The small groups that can organize or collude in the unstable societies will have different interests in different countries at different times. In one period they might be landed oligarchs, in another manufacturing firms; in one country they might have a vested interest in exports, in another, in import substitution. Again, it is important to respect the diversity and detail of actual experience and not to push this theory, or any other general theory, farther than it can go. The only general point so far about the unstable societies is that one must look at the vested interests of the small groups capable of relatively prompt collective action to understand one systematic element in economic policies.

By bringing additional information to bear we can make more specific predictions. One fact is that, since almost all the unstable governments are in developing nations, they usually do not have anything like a complete modern system of transportation and communication, at least in the rural areas. This makes it more costly and difficult for those in rural areas to mobilize political power to influence the government and gives the residents of the major metropolitan areas, especially the capital city, a disproportionate influence. Before the Industrial Revolution and the railway, transportation was slow and expensive everywhere, and this presumably explains why in his time Adam Smith observed that farmers were unable to organize to gain monopoly or political influence, whereas businessmen rarely met without conspiring against the public interest. In the developing nations today the rural interests are at a similar disadvantage, and the residents of the capital city obtain an altogether disproportionate share of governmental favors.

Although this was not always true, in most developing nations the largest firms and the wealthiest individuals are involved in producing import substitutes and goods that can also be provided by foreign firms. That is, they produce goods and services that are also available at lower cost on the world market or that could be provided more economically by local branches of foreign firms. The enterprises engaged in import substitution will in some developing countries include heavy industry, but in others they may only manufacture textiles or brew beer. Sometimes the wealthiest families will own banks or insurance companies that provide services that foreign firms could also provide. Of course, nowadays these enterprises and families will also tend to be located in the great cities with the easiest access to the government.

When the most substantial firms are in the import-substitution and foreign-replacement sectors, and especially when there is poor transportation in the rural areas, a special perverse policy syndrome develops in the ‘'top-heavy” society. The key to this policy syndrome is the markedly disproportionate strength of small groups in the unstable societies.

The large enterprises and wealthy families in the situations described above have an obvious interest in protection against imports and discriminatory legislation against the foreign or multinational firms with which they compete. This drives up prices for consumers, but consumers are in latent groups that cannot organize, and many of them are out in the provinces as well. The most drastic forms of protection, such as quotas and exchange controls that deny citizens foreign currency for the purchase of imports for which there are domestic substitutes, are often used. These methods of protection are not readily subject to measurement. But in many cases, the level of protection is staggeringly high and quite beyond comparison with the levels of protection in the major developed democracies.

Let us look at the effects of these protective policies on the distribution of income. The imports and foreign firms are normally a source of competition because they have lower costs. In other words, the unstable developing country and its firms do not have a comparative advantage in the types of production in question. If they did, in most cases they would not care for protection.²² Most developing countries do not have comparative advantage in many manufactured import substitutes because these types of production involve large proportions of capital and technical expertise, which are usually scarce and therefore tend to be relatively expensive.

When the goods and services that are intensive in capital and technical expertise are protected, the price of capital and technical expertise in the developing nation rises, particularly in the favored firms and industries. Some of the gains to owners of the favored enterprises will be consumed by the efforts to secure or maintain the political favors. The employees of some of these firms also may be able to share in the gains. Nonetheless, in at least some countries, the owners of capital and technical expertise, who were probably well rewarded because of their scarcity in the first place, now get even higher rates of return. Since it was the wealthier individuals and larger firms that initially were able to organize fastest, and since the mobilization of large amounts of capital and the acquisition of rare expertise requires wealth, the protection in most cases presumably favors the wealthy.

In addition, the protection makes the country’s currency more valuable than it would otherwise be; less of the national currency is supplied to buy the foreign exchange needed for imports. The higher the price of the national currency, the more expensive the country’s exports and the less of them foreigners will buy. The exports are in general the goods that the country has a comparative advantage in producing. Poor developing nations naturally have a lot of poor people and thus cheap labor, and some natural resources, so they tend to have a comparative advantage in producing goods that are intensive in labor and natural resources. The owners of labor and of natural resources, and the peasantry in particular, are the victims of the loss in exports. In some African countries, especially, the rural exporters are further exploited by government marketing monopolies that give the farmers only a portion of the price at which the government sells their commodities. The plentiful factors of production earn relatively little to begin with and the loss of export earnings reduces their incomes still further. The owners of the plentiful factors—which include in every developing nation the working poor—not only are denied access to cheaper imports by the protection, but also get lower prices for the labor and the products they have to sell.

There are a host of qualifications and technical niceties that it would be interesting to explore at this point. It is also important to point out that situations are somewhat different in each country. Nonetheless, the general nature of the process is clear. The most substantial and wealthy interests are relatively better organized in the unstable society, but they often own an unrepresentative mix of the country’s productive factors. They obtain policies that favor themselves and work in different ways against the interests of the larger unorganized groups in the society, thereby making the distribution of income far more unequal.

The available statistics are poor and incomplete, but it is clear that many of the unstable countries have unusually unequal distributions of income, with giant fortunes juxtaposed with mass poverty. In some of these countries it is obvious even from casual observation that, as the foregoing argument would lead us to expect, the unskilled workers are trying desperately to get out, and (when they think the chances of being nationalized are not great) the multinationals are trying to get in. A research assistant and I have compared some (very shaky) data on income distribution across countries with some (even shakier) data on the degree of instability. The weakness of the data and the multiplicity of alternative explanations of the results have forced me to conclude that the tests are not worth relating here. But for whatever little they are worth, they are consistent with the theory.

All the arguments and evidence in the preceding chapter about the losses from protection also apply to the unstable developing countries, especially the smaller ones. Thus the perverse policy syndrome described above promotes inefficiency and stagnation as well as inequality.

When the import-substituting industries in the capital and other metropolitan areas are protected and cartelized, they can survive even if they pay wages far above the competitive level. This allows greater gains from the monopolization of the labor force and promotes unions, although episodic repression and the difficulties of organizing large-scale collective action may prevent unionization. Whether unions emerge or not, the population of the capital city—even the poor—will tend to have more influence on public policy than their rural and provincial cousins because of the inadequacies of the transportation system. Popular demonstrations, strikes, and riots in the capital are a special threat to governments.

The civil and military bureaucracies, which are well placed to influence any government, will be disproportionately in the capital city. University students, with their untypically intense interest in politics and flexible schedules, are usually important in politics and are often in the large cities. The unions, the bureaucracy, and the students frequently will have different ideological colorations than the owners of large firms, but they may be equally disposed to economic nationalism and may not in practice be much in conflict over the detailed and inconspicuous policies the separate small groups normally seek. In any case, the bureaucracy, and often also the unions and the students, will support the subsidization of life in the capital and perhaps other large cities.

The provision of extra facilities and other forms of subsidization to urban areas encourages more migration to the capital and to other cities, beyond that already spurred by the import-substitution policies. So another aspect of the perverse policy syndrome is inefficiently large capital cities and major metropolitan areas. The capital cities of most of the poorer nations today are vastly larger in relation to the population of these nations than were the capital cities in the developed nations when those nations had the levels of per capita income the poorer countries have.

The unstable countries are so diverse, and their policies influenced by so many factors, that the preceding argument should be regarded as a “researchers’ parable” rather than an analysis; it should be read as a story meant to have heuristic value. Researchers and others with a specialized knowledge of particular unstable countries will, it is hoped, be stimulated to analyze the situation in a particular country, or some small set of countries, in a systematic way.

One type of inquiry that is needed is historical. Some of the countries that now have the perverse policy syndrome, or something rather like it, were probably once in a different situation. For some Latin American countries in the nineteenth century, for example, it is worth asking whether the group that could best organize to influence the government would have been a small group of landed families of great wealth. The difficulties of transportation would have made their collusion more difficult, and sometimes their power may have been exercised only locally, in a feudal fashion. If these families could collude on national policy, however, they would have had interests different from those who profit from the perverse policy syndrome. As owners of land and sometimes of labor in peonage, they would have held a mix of factors representative of the economies of which they were a part. As such they would have gained from liberal trade policies. To the extent that their poorer compatriots also owned some land and labor, they too would have profited. With revolutions and the gradual growth of cities, landed magnates lost influence or disappeared. Liberal and socialist writers in the capital cities pressed for egalitarian policies that did not appeal to the great landowners. There is a need for research to determine whether the changes have, in fact, reduced inequality. In some countries we may find that the small groups in urban areas that have influenced policy more recently have a vested interest in less efficient policies, and ones that have more inegalitarian consequences as well.

Now that we have considered the unique inequality of the caste system, the racial discrimination in South Africa, and the inegalitarian policies of the top-heavy societies with the perverse policy syndrome, we can examine a modern myth that has in my judgment forced needless poverty and humiliation on millions of people. Among economists, who are about the only people who have given the matter specialized study, there is a consensus that competitive markets are efficient. Indeed, in economic theory the definition of a perfectly competitive market entails that it is perfectly efficient, with the only possible improvement being in the distribution of income that results. Even among those who are in other vocations, there is, at least in the developed countries, a fairly widespread understanding that competition encourages efficiency.

There is at the same time the standard assumption, among economists as well as the laity, that competitive markets generate a considerable degree of inequality. A soft-hearted majority holds the further view that government action—or in some versions the operation of unions, professional ethics, and so forth—is needed to reduce the inequalities generated by the market. A hard-boiled minority willingly accepts or even rejoices in the inequality or believes that governmental efforts to reduce inequality are harmful. Most of the soft-hearted and most of the hard-boiled agree in taking it for granted that markets generate considerable inequality and differ about whether this inequality is unjust. The economist often speaks of the trade-off between efficiency, obtained through competitive markets, and equity, obtained at some social cost by other means.

Perhaps the most intelligent and humane expression of the view that competitive markets are a source of considerable inequality that government and other nonmarket institutions then reduce at some social cost is Arthur Okun’s widely respected book, Equality and Efficiency: The Big Trade-Off.²³ In this book it is taken for granted that governments are an egalitarian force that evens out the inequality resulting from the operation of markets, and that some price must be paid for this reduction in inequality because it interferes with the operation of generally efficient markets. Some other writers suppose that unions and other special-interest groups reduce the inequalities that result from competitive markets.

I submit that the orthodox assumption of both Left and Right that the market generates more inequality than the government and the other institutions that ‘'mitigate” its effects is the opposite of the truth for many societies, and only a half-truth for the rest. In South Africa there are black workers who are paid one-eleventh as much as white workers doing slightly different jobs that require about the same degree of skill. In a truly competitive economy, as the textbooks lucidly explain, it is difficult indeed to see how people with the same skills and effectiveness could earn very different rates of pay in the long run. Employers could profit by hiring the low-wage victims of discrimination, and firms that refused to do so would eventually be driven out of business by their lower-cost competitors. As we should expect, employers in South Africa argue that they should be allowed to use the African workers for jobs that are now restricted to whites. If this is allowed, both efficiency and equity will improve. In India many tens of millions of people historically have been condemned, and their children after them, to lives of special poverty and humiliation by caste rules that prevent the free operation of markets; and these rules have led to inefficiency and stagnation as well as inequality. Most of the countries of this world are unstable developing nations, and in most of them the policies on international trade, foreign investment, and many other matters make these societies generate colossal inequalities as well as inefficiency. This is evidence that, as I argued in justification of Implication 9, the gang fight is no gentler than the individual duel.

Now let us turn to the developed democracies with their welfare states. Those of us who believe that we ought to make a decent provision for the least fortunate in our societies, even though it will require that we make some sacrifices ourselves, have to face up to the logic of collective action. We can help our friends, relatives, or neighbors and can see the benefits that result from our generosity, so we may make significant sacrifices on their behalf. But if we strive as individuals to reduce the poverty in the country in which we live, we find that even if we gave up all of our wealth it would not be enough to make a noticeable difference in the amount of poverty in the society. The alleviation of poverty in a society is, in other words, a public good to all those who would like to see it eliminated, and voluntary contributions will not obtain public goods for large groups. If everyone who is concerned about poverty made a contribution to its alleviation, in the aggregate that would, however, make a difference. So a majority of us in each of the developed democracies votes for raising some money by imposing compulsory taxes on ourselves, or more precisely on the whole society, and devoting these monies to the needy. Since the alleviation of poverty on a society-wide basis is a public good, efforts to redistribute income to the poor as a group require governmental action. In this respect, it is true that governments in some societies do mitigate inequalities. Since both the taxes and the transfers have adverse effects on incentives, it is also true that there are trade-offs between equality and efficiency.

The trouble is that the current orthodoxies of both Left and Right assume that almost all the redistribution of income that occurs is the redistribution inspired by egalitarian motives, and that goes from the nonpoor to the poor. In reality many, if not most, of the redistributions are inspired by entirely different motives, and most of them have arbitrary rather than egalitarian impacts on the distribution of income— more than a few redistribute income from lower to higher income people. A very large part of the activities of governments, even in the developed democracies, is of no special help to the poor and many of these activities actually harm them. In the United States there are subsidies to the owners of private airplanes and yachts, most of whom are not poor. The intervention of the professions and the government in the medical care system, as I have shown elsewhere,²⁴ mainly helps physicians and other providers, most of whom are well-heeled. There are innumerable tax loopholes that help the rich but are without relevance for the poor and bailouts for corporations and protection for industries when the workers’ wages are far above the average for American industry. There are minimum-wage laws and union-wage scales that keep employers and workers from making employment contracts at lower wages, with the result that progressively larger proportions of the American population are not employed. The situation in many European countries is much the same and in some cases a little worse.

The reason that government and other institutions that intervene in markets are not in general any less inegalitarian than competitive markets is evident from the discussion in chapter 3 of Implications 3 and 9. There is greater inequality, I hypothesize, in the opportunity to create distributional coalitions than there is in the inherent productive abilities of people. The recipients of welfare in the United States are not organized, nor are the poor in other societies. But in the United States, as elsewhere, almost all the major firms are represented by trade associations and the professions by professional associations. There are admittedly differences in the productive abilities of individuals, just as there are differences in height. But such measurement as we are now capable of suggests that the individual differences are normally distributed—the vast majority at least fairly close to the middle. There are a few dwarfs and a few giants, but not many. Larger differences are apparent, it is true, in the holdings of capital and some huge fortunes. Yet, if the accumulation of capital is unobstructed and policies such as those in the perverse policy syndrome are avoided, the return to capital will fall as more capital is accumulated²⁵ and the wages of the labor with which the capital is combined will rise; it is no accident that wages are highest in the countries that have enjoyed the greatest accumulation of capital. If economic nationalism does not keep the capital from crossing national borders, more will have an incentive to migrate to areas with the lowest wages and thus significantly raise the wages of the poorest. As the history of India tells us, even in the longest run there is no comparable tendency for inequalities to diminish over time through distributional coalitions.

XII

Another myth that generates a lot of poverty and suffering is that the economic development of the poor countries is, for fundamental economic or extra-institutional reasons, extremely difficult, and requires special promotion, planning, and effort. It is sometimes even argued that a tough dictator or totalitarian repression is required to force the sacrifices needed to bring about economic development. As I see it, in these days it takes an enormous amount of stupid policies or bad or unstable institutions to prevent economic development. Unfortunately, growth-retarding regimes, policies, and institutions are the rule rather than the exception, and the majority of the world’s population lives in poverty.

The examples of successful growth that have been referred to in this study did not occur because of any special promotion or plans. Neither did that of Korea, Taiwan, Hong Kong, or Singapore. The former two received some aid from the United States, but they also felt compelled to spend unusually large amounts on military purposes. If the analysis in this book is right, the growth in Germany and the United States before World War I was more the result of a widening of product and factor markets than of any special promotion or plan. So it was with the growth in early modern Europe. Britain did not seek or plan to have an industrial revolution; it grew for other reasons such as those explained earlier in this book. In the many countries that have failed to grow or failed to grow as fast or as far as the leaders, there are quite enough stupidities, rigidities, and instabilities to explain the lack of success.

Some people suppose that it is more difficult for poor nations to grow now than it was in the eighteenth or nineteenth centuries, and that the explanation for this is in some sense also economic rather than institutional. This overlooks the fact that the poor nations now can borrow the technologies of more developed nations, some of which will be readily adaptable to their own environments, and improve their techniques of production very rapidly. Great Britain in the Industrial Revolution could improve its technology only through the inventions that occurred in that period. Similarly, most highly developed nations today can improve their technology only by taking advantage of current advances. The poorest of the developing nations can telescope the cumulative technological progress of centuries into a few decades. This is not only an obvious possibility but has actually occurred in places like Korea, Taiwan, Hong Kong, and Singapore. The nations of continental Europe and Japan were far behind the United States technologically at the end of World War II, but they borrowed American technology, grew far faster than the United States, and very nearly caught up with the United States in both technology and per capita income in less than twenty-five years.

In claiming that international product and factor markets unobstructed by either cartelization or governmental intervention will bring irrepressible and rapid growth to any poor country, I am not arguing that laissez-faire leads to perfect efficiency. As I pointed out in chapter 3, 1 do not assume perfect competition anywhere in this book. As it happens, most of my own writing in economics is about externalities and public goods, which normally keep a laissez-faire economy from achieving Pareto-optimality and which I believe are quite important.²⁶ An economy can be dynamic and rapidly growing without at the same time being optimal or perfectly efficient. An economy with free markets and no government or cartel intervention is like a teen-aged youth; it makes a lot of mistakes but nonetheless grows rapidly without special effort or encouragement.

If poor institutions that prevent or repress growth are the norm in much of the world, it may not help to say that “only” institutional problems stand in the way of rapid growth in poor countries. If poor institutions are so common, it presumably is not always easy to obtain good institutional arrangements and the rapid growth that they permit. Still, the problems are more likely to be solved if they are understood than if they are not.²⁷

XIII

As might be expected from my concern about ideological preconceptions, and from the methodological discussion in chapter 1 on the standards a satisfactory answer must meet, I do not believe any of the ideological approaches are sufficient to meet our needs. In keeping with that belief, I want to underline the contrast between the argument here and the classical liberal or laissez-faire ideology. The present argument and the classical liberal ideology do share an appreciation of the value of markets. An appreciation of markets is common to nearly everyone, Right or Left, who has given the matter a decade or more of specialized study. If you stand on the shoulders of the giants, it is virtually impossible to see it any other way.

But there the similarity between the present argument and classical liberal laissez-faire ideology stops. As I read it, the ark and covenant of the laissez-faire ideology is that the government that governs least governs best; markets will solve the problem if the government only leaves them alone. There is in the most popular presentations of this ideology a monodiabolism, and the government is the devil. If this devil is kept in chains, there is an almost Utopian lack of concern about other problems.

If the less optimistic theory in this book is right, there often will not be competitive markets even if the government does not intervene. The government is by no means the only source of coercion or social pressure in society. There will be cartelization of many markets even if the government does not help. Eliminating certain types of government intervention and freeing trade and factor mobility will weaken cartels but will not eliminate many of them. Moreover, the absence of government intervention (even if it were invariably desirable) may not be possible anyway, because of the lobbying of special-interest groups, unless we fly to the still greater evil of continuous instability.

The questions of whether laissez-faire alone is sufficient to prevent or eliminate cartelization, and whether laissez-faire is in the long run not viable because special-interest groups will accumulate and lobby it out of existence can be settled only by an appeal to the facts. Thanks to British imperialism, history has given us one experiment of remarkable aptness. Milton and Rose Friedman, in Free to Choose,²⁸ made much of a comparison between Japan after the Meiji restoration in the late 1860s and India since World War II. The point of comparison was, of course, that the Japanese after the Meiji restoration had relatively free enterprise, along with very low tariffs, whereas independent India has had dramatically interventionist and protectionist policies. As the Fried-mans correctly point out, the policies the Japanese chose produced great growth, but those India chose failed. There is a great deal to be said for this comparison and for the policy lesson the Friedmans draw from it. I should also point out that, like most other economists of my generation, I have learned a lot from Milton Friedman’s exceptionally lucid, fresh, and penetrating technical writings.²⁹ And I greatly respect the depth of the Friedmans’ convictions.

Withal, there is an ideological—as opposed to a scientific—element in the comparison, and an instructive one at that. Of their comparison, the Friedmans write, “Economists and social scientists in general can seldom conduct controlled experiments of the kind that are so important in the physical sciences. However, experience here has produced something very close to a controlled experiment to test the importance of the difference in methods of economic organization. There is a lapse of eight decades in time. In all other respects the two countries were in very similar circumstances.”³⁰

An even closer approximation to the controlled experiment of the physical sciences is possible—the same one, without the “lapse of eight decades in time.” It is all too often forgotten that one of the finest examples of laissez-faire policy was British rule in India. India had one of the most thoroughgoing laissez-faire policies the world has seen, and it was administered with considerable economy and efficiency in the best British civil service tradition.³¹ Entrepreneurs and capitalists from all over the world were free to sell or buy in India or to set up businesses there, as were Indians themselves. No doubt there must have been favoritism to British firms, but where and when in human history was there much more laissez-faire impartiality? There was less government intervention than in Japan after the Meiji restoration. Tariffs were used only to raise revenue and part of the time there were not even revenue tariffs. Those who incorrectly ascribe most economic development to state intervention might claim that India failed to grow because it did not have an independent government that could engage in economic planning and promote development. This argument is not, however, open to the advocate of laissez-faire ideology, for that ideology does not require an active independent government, and in any case the experience of Hong Kong argues that colonies can grow with extraordinary rapidity.³²

A half-century or more of laissez-faire did generate some growth in India, but nothing comparable to what occurred in Japan. Laissez-faire led to some change and loosening of India’s caste system, and some new special-interest organizations emerged. My guess is that if India after World War II had followed the policies the British once required, it would have done better than it has. Nonetheless, the fact remains that more than a half-century of laissez-faire did not bring about the development of India or even get it off to a good start. The laissez-faire ideology in its focus on the evils of government alone clearly leaves something out. I submit that it is the distributional coalitions, which over millennia of history in India had hardened into castes.

Another great experiment in laissez-faire was conducted in Great Britain itself. Britain generally followed laissez-faire policies at home as well as abroad from about the middle of the nineteenth century until the interwar period. (The United States in the same period had highly protectionist policies and in this and some other respects, such as the subsidies to railroads, fell considerably short of laissez-faire.) In this book I have argued, as would the ideological enthusiasts for laissez-faire, that the free trade policy Britain followed has limited the extent of distributional coalitions there. Things could be worse in Britain, and would have been had Britain had the highly protectionist policies of Australia and New Zealand. Nonetheless, as the theory here argues, laissez-faire did not prove to be dynamically stable—Britain abandoned it. Neither was it sufficient to prevent cartelization in many sectors.

During the nineteenth and early twentieth centuries, precisely when and where laissez-faire policy was at its peak, Great Britain acquired a large proportion of its dense network of narrow distributional coalitions. It was in this same period, too, that the British disease emerged and British growth rates and income levels began to lag.

II

III

IV

V

VI

VII

VIII

IX

X

XI

XII

XIII