The index of social development is the backbone of this book, holding together the body of facts that archaeologists and historians have accumulated. The index does not itself explain why the West rules, but it does show us the shape of the history that has to be explained. I provide a full account of the index, for those interested in the methods and detailed evidence behind the calculations, at the website www.ianmorris.org; this appendix is intended only as a quick summary of the main technical challenges and the basic results.
FOUR OBJECTIONS
I see four obvious objections to the social development index:
1. Quantifying and comparing social development in different times and places dehumanizes people and we should therefore not do it.
2. Quantifying and comparing societies is a reasonable procedure, but social development in the sense I defined it (as societies’ abilities to get things done) is the wrong thing to measure.
3. Social development in the sense I defined it is a useful way to compare East and West, but the four traits I used to measure it (energy capture, organization/urbanization, war-making, and information technology) are not the best ones.
4. These four traits are a good way to measure social development but I have made factual errors and got the measurements wrong.
I addressed objection 1 in Chapter 3. There are plenty of historical and anthropological questions for which quantifying and comparing social development is no help at all, but asking why the West rules is by its nature a comparative and quantitative question. If we want to answer it, we must quantify and compare.
I also said a few words in Chapter 3 about objection 2. Perhaps there are other things we could measure and compare that would work better than social development, but I do not know what they are. I leave it to other historians and anthropologists to identify other objects to measure and to show that they yield better results.
Objection 3 can take three forms—that we should add more traits to my four; that we should use different traits; or that we should look at fewer traits. As I wrote this book I did explore several other traits (for example, area of largest political unit, standards of living [measured through adult stature], transportation speeds, or size of largest monuments), but all had severe evidence problems or failed the test of mutual independence. Most traits in any case show high levels of redundancy through most of history, and any plausible combination of traits will tend to produce much the same final result.
There are plenty of small and two large exceptions to the redundancy rule. The first large exception is what we might call the “nomad anomaly”—the fact that steppe societies normally score low on energy capture, organization, and information technology, but high on war-making. This anomaly helps explain why true nomad societies have been so good at defeating empires but so bad at running them,* and it deserves extensive study, but it does not directly affect the comparisons between the settled, agrarian Eastern and Western cores in this book.
Figure A.1. Energy alone: how East and West compare if we just look at energy capture per person
Another version of objection 3 would drop organization, war-making, and information technology from the analysis and concentrate only on energy capture, on the grounds that organization, war-making, and information technology are merely ways of usingenergy. Figure A.1 shows what an energy-alone index would look like. It is different from the full index graph in Figure 3.3, but not hugely so. In the energy-alone graph, just like the full social development graph, the West still leads the East for 90 percent of the time, the East still overtakes it between roughly 550 and 1750 CE, there is still a hard ceiling that blocks development around 100 and 1100 CE (at just over 30,000 kilocalories per person per day), postindustrial revolution scores still dwarf those of earlier ages, and in 2000 the West still rules.
Focusing on energy alone has the advantage of being more parsimonious than my four-trait approach to social development, but it also has one great drawback. This is the second large exception to the redundancy rule: the fact that since the industrial revolution the relationship between traits has become nonlinear. Thanks to new technologies, city size quadrupled across the twentieth century, war-making capacity increased fiftyfold, and information technology surged eightyfold, while energy capture per person merely doubled. Looking at energy alone is too simple, and distorts the shape of history.
Objection 4 raises very different issues, because the only way to assess whether I have misunderstood the evidence or used inappropriate methods is by reexamining all the sources of information I used to calculate Eastern and Western scores across the last sixteen thousand years. Doing that in this appendix would be an expensive proposition, making an already-long book much longer still, so I have put the information on the website mentioned above. Readers with the time and inclination can find out there precisely what sources I have used and my views on the ambiguities in the evidence.
In what remains of this appendix I will summarize the data, outline quickly how I calculated the scores, and say a few words about margins of error.
ENERGY CAPTURE
I discuss energy capture first and at greatest length because it is quantitatively the most important of the four traits. If we go back far enough in time, the urbanization, war-making, or information-technology scores all fall to zero because human activities were on such a tiny scale that they generate values below 0.01 point on the index. The energy-capture scores by contrast never fall to zero, because humans who capture zero energy die. Keeping body and soul together requires roughly 2,000 kilocalories per capita per day, and since modern Western energy capture is about 228,000 kcal/cap/day (= 250 points), the lowest score possible in theory would be 2.19; and in reality, energy capture has always scored above 4 points since the end of the Ice Age, because much of the energy humans use is in nonfood forms (clothes, shelter, artifacts, fuel, and so on). Until the industrial revolution, the energy capture score typically accounts for 75–90 percent of the total social development scores. In 2000 it still accounted for 28 percent of the Western and 20 percent of the Eastern scores.
The evidence for energy capture ranges from modern statistical digests to literary accounts of farming, industry, and lifestyles, to archaeological evidence for diet, crafts, and quality of life. Combining such varied materials is a challenge, but here, as elsewhere, I have built on the contributions of earlier researchers. As I explained in Chapter 3, Earl Cook’s 1971 study of energy flows provides a convenient starting point that can be constantly checked against other estimates. These all converge on contemporary levels in the Western core of around 230,000 kcal/cap/day, which Cook divides into rough categories of feed/food (for domesticated animals as well as humans), home/commerce, industry/agriculture, and transport.
Vaclav Smil (1991, 1994) usefully breaks nonfood consumption down into biomass and fossil fuels, and graphs their development in the Western core over time. Several steps are needed to turn his data into energy-capture scores for the West, but the results come out around 93,000 kcal/cap/day in 1900 and 38,000 in 1800, neatly bracketing Cook’s estimate of 77,000 for industrialized Europe in 1860.
The further we move back before 1800 the fewer government-generated statistics are available, but the more that economies relied on biomass fuels, the more we can substitute comparative information gathered by economic historians and anthropologists for official documents. In 1700 the average person in the Western core must have consumed somewhere between 30,000 and 35,000 kcal/day. Our evidence for what Western societies did shows clearly that the further we go back into the previous thousand years the lower that number falls,* though the comparative evidence also makes it clear that Western energy consumption could never have fallen too far below 30,000 kcal/cap/day. There is room for debate, but I doubt that medieval Western energy capture ever fell below 25,000 kcal/cap/day, even in the eighth century CE. For reasons I return to below, I do not see how these guesstimates can be more than 5–10 percent wide of the mark.
Table A.1. Energy capture, kilocalories/person/day (selected dates)
The impressive ruins of Roman-era houses and monuments, the numbers of shipwrecks, the volume of manufactured goods, the level of industrial pollution in ice cores, and the staggering numbers of animal bones from settlements make it clear that Western energy capture was higher in the first century CE than in the eighth or even the thirteenth, but how much higher? Ingenious calculations by economic historians point toward an answer. Robert Allen (2007a) has shown that in 300 CE real wages (which, for most of the poor in premodern times, closely mirrored energy consumption) in the Western core were comparable to those of southern Europe in the eighteenth century CE, and Walter Scheidel (2008) has suggested that Roman-era wages were comfortably higher than those in much of medieval Europe. Data gathered by Geof Kron (2005) and Nikola Koepke and Joerg Baten (2005, 2008) indicate that stature changed little between the first and eighteenth centuries, and Kron (forthcoming) suggests that ancient housing was typically better than that in the richest parts of eighteenth-century Europe. I have estimated energy capture at around 31,000 kcal/person/day in the years 1 BCE/CE, declining slowly until 500 CE and then faster until 700.
Energy capture must have been lower in the Western core around 1000 BCE not only than in Roman times, but also than in the eighth century CE. The sharpest period of increase came after 300 BCE, as the Mediterranean was integrated into larger political and economic units and the Roman Warm Period raised output, but the mass of archaeological data also shows an earlier period of acceleration after 600 BCE. I have tentatively suggested that in 1000 BCE energy capture may have been as low as 20,000 kcal/cap/day, a slight decline on the levels of the late second millennium BCE, but still above those of the third millennium.
Earlier in prehistory scores were lower still. At the end of the Younger Dryas foragers were probably getting by on about 5,000 kcal/cap/day, but this would have risen sharply (relative to what had gone before) as the climate warmed, plants and animals were domesticated for food, and animals were harnessed for draft power. By 3000 BCE people in established villages in the Hilly Flanks must have been consuming 12,000 kcal/cap/day for their clothes, fuel, farm animals, houses and household goods, and monuments, even if their diets were no better than they had been four millennia earlier.
Calculating Eastern scores is more difficult still, partly because scholars such as Cook and Smil were concerned only with the region of the world that had the highest energy capture, not with regional comparisons. We can begin, though, from the United Nations (2006) estimate that in 2000 CE the average Japanese person consumed 104,000 kilocalories per day (less than half the Western level). In 1900 the Eastern core was still largely agrarian, with Japanese oil use and even coal-powered industry in its infancy. Japanese energy capture may have been around 49,000 kcal/cap/day (again less than half of Western consumption). Across the previous five centuries coal use and agricultural output had risen steadily. In 1600 productivity was higher in the Yangzi Delta than anywhere in the West, but by 1750 Dutch and English agriculture had caught up and Eastern real wages were comparable to those in southern Europe rather than wealthy northern Europe. I have estimated energy capture in the Eastern core around 29,000 kcal/cap/day in 1400 and 36,000 in 1800, with the bulk of the increase coming in the eighteenth century.
There is also debate over how badly the crisis after 1200 impacted Chinese energy use, but there was probably at least a slight dip from the Song-era peak, when consumption probably surpassed 30,000 kcal/cap/day.
As in the West, the archaeological evidence makes it clear that energy capture went through a trough in the mid first millennium CE, but again it is difficult to say just how steep the decline was. The evidence I reviewed in Chapter 5 suggests that Han dynasty energy consumption was higher than anything previously seen in the East but lower than contemporary Roman or later Song levels; I have estimated 27,000 kcal/cap/day in 1 BCE/CE, returning to the same level by 700 CE after a slight fall.
Again paralleling the West, Eastern energy capture in the first millennium BCE saw steady increases, accelerating after about 500 BCE and still more after 300 BCE with the spread of canal networks, trade, and metal tools. Back in 1000 BCE the average energy capture may have been around 17,000 kcal/cap/day; by the time of the Qin First Emperor it was probably more like 26,000.
In prehistoric times Eastern energy capture seems to have passed through much the same thresholds as Western, but began moving upward later and generally ran one to two millennia behind.
ORGANIZATION
Throughout preindustrial history organization was always the second-largest component in social development scores. I used this trait as my main example in Chapter 3, explaining why I use largest city size as a proxy for social organization. There is enough ambiguity in the data and flexibility in definitions that experts disagree over city sizes in every period, and I explain my decisions on the website. In Table A.2 I just summarize some of my main calculations.
WAR-MAKING
Since writing began, people have recorded their wars, and since early prehistory have often buried their dead with weapons. As a result we know a surprising amount even about premodern warfare. The major challenge in scoring war-making capacity is not empirical but conceptual—how we compare radically different fighting systems that are often intended to be incomparable with earlier systems. Most famously, when Britain launched HMS Dreadnought in 1906, the whole idea was that its supersized guns and heavy armor meant that no number of 1890-era ships would add up to one post-1906 ship.
The reality, though, is that things never work out so simply. Improvised explosive devices can, under the right circumstances, give even the highest-tech army a run for its money. In principle we can assign scores on a single index to wildly different military systems, even if experts might argue over what those scores should be.
In 2000 CE, the West’s unprecedented military power earns 250 points, and is clearly much greater than the East’s. Some Eastern armies are large, but weapons systems matter far more than sheer numbers. The United States outspends China 10:1, and outnumbers it 11:0 in carrier groups and 26:1 in nuclear warheads. The qualitative differences between America’s M1 battle tanks and precision weapons and China’s outdated systems are still greater. Setting the West:East points ratio as low as 10:1 or as high as 50:1 both seem extreme, and I have guessed at 20:1, meaning that the East scores 12.5 points in 2000 as against the West’s 250.
Table A.2. Population of the largest settlement in each core, thousands (selected dates)
Comparing scores in 2000 with those in earlier periods is even more difficult, but by looking at the changes in the size of forces, the speed of their movement, their logistical capacities, the range and destructiveness of their striking power, and the armor and fortifications at their disposal, we can make rough estimates. According to one calculation, the effectiveness of artillery fire increased twentyfold between 1900 and 2000 and that of antitank fire sixtyfold; factoring in all the other changes across the twentieth century, I set the ratio between Western war-making capacity in 2000 and 1900 at 50:1, meaning that the West scores 5 points in 1900 compared with its 250 points in 2000.
Western military power in 1900 was much greater than Eastern, though the gap was certainly not as large as it was by 2000. The British navy had nearly six times the tonnage of the Japanese in 1902, and any one of Europe’s great powers had more men under arms than Japan; I set the West:East ratio in 1900 at 5:1, meaning that the East scores just 1 point in 1900 (as compared with the West’s 5 points in 1900 and the East’s 12.5 points in 2000).
Not everyone will be comfortable with the level of subjectivity in calculations such as these, but the important point is that the West’s military capacity in 2000 was so enormous that all other scores—including the West’s in 1900 or even the East’s in 2000—are necessarily tiny; and, as a result of this, the errors involved in estimation are insignificant. We could double, or cut in half, any or all of the war-making scores for all periods up to 1900 without having a noticeable impact on the total development scores.
The contrast between Western war-making in 1800 and that in 1900 was less than the contrast between 1900 and 2000, but it was still enormous, taking us from the age of sail, cavalry charges, and smoothbore muzzle-loaded muskets to that of explosive shells, armored oil-powered ships, and machine guns, with tanks and aircraft just around the corner. The nineteenth century probably raised Western war-making capacity by an order of magnitude, and I set the West’s war-making capacity at just 0.5 point in 1800. Western warfare at that point was vastly more effective than Eastern, which should perhaps earn just 0.1 point in 1800.
Between 1500 and 1800 Europe went through what historians commonly call a “military revolution,” perhaps quadrupling the effectiveness of its war-making. Eastern war-making, by contrast, actually went backward between 1700 (when Kangxi began conquering the steppes) and 1800. In the absence of existential threats, Chinese rulers regularly sought peace dividends by reducing their armed forces and ignoring expensive technological advances. Eastern war-making was not noticeably more effective in 1800 than it had been in 1500, and was much less effective than it had been in 1700—which has a lot to do with why Britain’s forces swept China’s aside so easily in the 1840s.
The advent of gunpowder weapons in the fourteenth century increased war-making capacity in both East and West, though much less dramatically than the inventions of the nineteenth and twentieth centuries would do. In Europe the best armies around 1500 (particularly the Ottomans) were probably twice as effective as those of five centuries earlier, though that had as much to do with size and logistics as with firepower.
The relationship between Western war-making around 1500 and the large, highly organized, but pre-gunpowder forces of the Roman Empire is harder to calculate. One study has estimated that a single jet bomber around 2000 CE had half a million times the killing capacity of a Roman legionary, which we might take as implying that the Western score in 1 BCE/CE would be 0.0005 point; but of course Rome had far more legionaries than the United States has jet bombers, and I estimate the ratio between modern Western and Roman war-making at more like 2,000:1, putting the Western score in 1 BCE/CE at 0.12 point. That makes the Roman war machine at its height a serious rival for fifteenth-century European armies and navies, despite their guns and cannons, but not for the forces of the “military revolution” era. It also means that Roman war-making at its zenith might have competed with that of the Mongols and was superior to that of Tang dynasty China.
In the East, where bronze weapons were still the norm as late as 200 BCE, Han dynasty (200 BCE–200 CE) forces seem to have been less effective than Roman, although Chinese military power declined much less than Western after the Old World Exchange. The armies and navies that the Sui used to reunite China in the sixth century were much stronger than anything in the West, and by the time of Empress Wu around 700 the gap was enormous.
The militaries of the centuries BCE were much weaker than those of the Roman and Han empires. In the East I assume that no force before the time of Erlitou around 1900 BCE was effective enough to score 0.01 point; in the West, Egyptian and Mesopotamian armies probably scored 0.01 point by about 3000 BCE.
Table A.3. War-making capacity, expressed in points on the social development index (selected dates)
INFORMATION TECHNOLOGY
Archaeological and written sources show us what kinds of information technology existed at various periods and it is not too difficult to estimate how much information these media could communicate, at what speeds, and over what distances. The real problem lies in estimating the extent of use of different technologies, which through most of history means how many people could read and write and at what levels of competence.
Moore’s Law—that the cost-effectiveness of information technology has doubled every eighteen months or so since about 1950—seems to imply that the score in 2000 should be about a billion times higher than that of 1900, giving us a Western score in 1900 of 0.00000025 point. But that, of course, would overlook both the flexibility of old-fashioned forms of information storage such as printed books (which digital media are only now beginning to challenge) and changes over time in access to the most advanced techniques.
The correct ratio between modern and earlier information technology is much less than a billion to one, though it is clearly enormous, with the consequence that pre-1900 scores (and even more so, pre-1900 margins of error) are even tinier than in the case of war-making. On the other hand, the evidence for just how many people could read, write, and count at various levels of skill is much vaguer than the evidence for war, and my guesstimates are even more impressionistic.
In table A.4 I take a multistep approach to quantifying information technology. First, following common practice among historians, I divide skills into full, medium, and basic. The bars for each category are set low—in terms of literacy, basic means being able to read and write a name; medium, being able to read or write a simple sentence; and full, being able to read and write more connected prose. The Chinese Communist Party’s definitions in its 1950 literacy drive (full literacy, being able to recognize 1,000 characters; semiliteracy, recognizing 500–1,000; basic, 300–500) were rather similar.
Second, drawing on the available scholarship, I divide the adult male population at different periods across these three categories. For each 1 percent of men that falls into the full-literacy category I assign 0.5 point; for each percent in the medium category, 0.25 point; and for each percent in the basic category, 0.15 point. I then assign the same scores to women. The evidence for female literacy is poorer than for male, though it is clear that until the twentieth century fewer (usually far fewer) women than men could read and write. Although I am basically guessing before recent times, I hazard estimates of female use of information technology as a percentage of male use. I then assign points to each period based on the amount and level of information technology use.
In 2000, 100 percent of men and women are in the full category in both the Western and Eastern cores,* scoring 100 information technology points for both regions. In 1900, nearly all men in the Western core had at least some literacy (50 percent full, 40 percent medium, and 7 percent basic) and women were almost as well educated, generating a Western score of 63.8 IT points. In the East literacy was also widespread among men, though not at such high levels (I estimate 15 percent full, 60 percent medium, and 10 percent basic), though literate women may have been only a quarter as common. The result is an Eastern score of 30 IT points. As I repeat these calculations further back in history, the possible margin of error around my guesses steadily increases, although the tiny numbers of literate people make the impact of these errors correspondingly small.
The third step is to apply a multiplier for the changing speed and reach of communication technologies. I divide the most advanced tools for handling information into three broad categories: electronic (in use in both East and West by 2000), electrical (in use in the West by 1900), and preelectrical (in use in the West for perhaps eleven thousand years and in the East for perhaps nine thousand years).
Unlike most historians I do not make a strong distinction between print and preprint eras; the main contribution of printing was to produce more and cheaper materials rather than to transform communication the way that the telegraph or Internet would do, and these quantitative changes have already been factored in. For electronic technologies, I use a multiplier of 2.5 for the West and 1.89 for the East, reflecting the relative availability of computers and broadband communication in West and East in the year 2000. For electrical technologies, having some impact in the West by 1900, I use a multiplier of 0.05; and for preelectrical technologies, in use in all other periods, I use a multiplier of 0.01 in both East and West. Consequently, in 2000 the West scores the maximum possible 250 social development points (100 IT points × 2.5) and the East gets 189 (100 IT points × 1.89); in 1900, the West scores 3.19 points (63.8 × 0.05) and the East 0.3 (30 × 0.01). The Western score reaches the minimum level necessary to register on the index of social development (that is, 0.01 point) only around 3300 BCE; the Eastern, around 1300 BCE.
Table A.4. Information technology scores
MARGINS OF ERROR
I spoke repeatedly of estimates and guesses in the previous section, because there is no way to build a social development index without them. One consequence of that is that no index will ever be “right,” whether we take that word in the strong sense of meaning that every single detail is completely accurate or in the weaker sense of meaning that all experts will make the same estimates. As a result, there is no point in asking whether the social development scores I have calculated are wrong. Of course they are. The real question is: How wrong are they? Are they so wrong that the basic shape of the history of social development as represented in the graphs in Chapters 4–10 is misleading, meaning that this whole book is fatally flawed? Or are the errors in fact relatively trivial?
These questions can in principle be answered easily enough; we simply need to ask (1) just how much we would need to change the scores to make the past look so different that the arguments advanced in this book would cease to hold good and (2) whether such changes are plausible.
Ultimately the only way to do this is by examining the evidence listed on the website (www.ianmorris.org) for each individual calculation I have made, but here I want to address briefly the possibility that systematic errors undermine my claims about the overall shape of history. According to my index (shown on a log-linear scale in Figure 3.7), the West took a lead after 14,000 BCE. The East slowly caught up, and through most of the first millennium BCE the West’s lead was narrow. Around 100 BCE the West increased its lead, but in 541 CE the East pulled ahead. It stayed there until 1773. The West then regained the lead, and, if twentieth-century trends continue, will hold it until 2103. Western development has been higher than Eastern for 92.5 percent of the time since the end of the Ice Age.
I suggested in Chapter 3 that overall my scores could err by as much as 10 percent without significantly altering this pattern. Figure A.2a shows how the trends would look if I have consistently underestimated Western development scores by 10 percent and overestimated Eastern scores by the same amount; Figure A.2b shows the outcome if I have underestimated Eastern development scores by 10 percent and overestimated Western scores by the same amount.
The first point to note is that these scores severely strain credibility. Figure A.2a, raising Western and lowering Eastern scores by 10 percent, requires us to accept that the West was more developed than the East in 1400 CE, right before Zheng He sailed on the Indian Ocean; it also means that when Hannibal led his elephants to attack Rome in 218 BCE Western development was already higher than the East’s would be in Zheng’s time. As if that were not peculiar enough, the graph additionally tells us that the West was more developed when Julius Caesar was murdered in 44 BCE than the East was when China’s emperor Qianlong rejected Lord Macartney’s trade embassy in 1793.
Figure A.2b is perhaps even more peculiar. The development score it gives to the West in 700 CE, for instance, when the Arabs ruled a vast caliphate from Damascus, is lower than that for the East in the age of Confucius, which cannot be right; and it would make the Western score in 1800, when the industrial revolution was already under way, lower than the Eastern scores under the Song dynasty in 1000–1200, which seems even less likely.
Yet even if historians could swallow such odd conclusions, the shapes of history as represented in Figure A.2 are still not different enough from that in Figure 3.7 to change the basic pattern that needs explaining. Short-term accident theories remain inadequate because even in Figure A.2b the West’s score is still higher most of the time (although “most” now means 56 percent rather than 92.5); so, too, long-term lock-in theories, because even in Figure A.2a the East does take the lead for seven centuries. Biology and sociology remain the most plausible explanations for the upward but interrupted movement of development, while geography remains the most plausible explanation for why the West rules.
Figure A.2. Error revealed: the implications of systematic errors in social development scores. (a) raises all Western scores by 10 percent and reduces all Eastern scores by the same amount; (b) raises all Eastern scores by 10 percent and reduces all Western scores by the same amount.
To change the fundamental patterns my estimates would need to be 20 percent wide of the mark. Figure A.3a shows how history would look if I have consistently underestimated Western development scores by 20 percent and overestimated Eastern scores by the same amount; Figure A.3b, the outcome if I have underestimated Eastern development scores by 20 percent and overestimated Western scores by the same amount.
This time the patterns are very different. In Figure A.3a the Western score is always higher than the Eastern, making long-term lock-in theories seem very plausible and also invalidating my claim that social development changes the meaning of geography.Figure A.3b, by contrast, effectively reverses the conclusions of my actual index, having the East lead 90 percent of the time since the Ice Age.
If either Figure A.3a or A.3b is correct, everything you have just read in this book is wrong. We can be confident, though, that they are not correct. Figure A.3a, raising Western scores and reducing Eastern scores by 20 percent, tells us that imperial Rome’s development in 1 BCE/CE was only five points behind industrial Japan’s in 1900, which cannot be true; while Figure A.3b, raising Eastern scores and reducing Western scores by 20 percent, means that Eastern development was higher in pre-Shang times than Western would be under the Persian Empire; that the West caught up with the East only in 1828, on the eve of the Opium War; and that Western rule has already ended (in 2003). None of this seems credible.
Hence my suggestions in Chapter 3 that (a) the margin of error in my estimates is probably less than 10 percent and definitely less than 20 percent, and (b), even if the margin of error does rise to 10 percent, the basic historical patterns I am trying to explain still hold good.
CONCLUSION
I observed several times in Chapter 3 that making a social development index is chainsaw art. The best an index can do is to give us a rough, good-enough approximation that makes the index designer’s assumptions explicit. I have argued that the main reason we have for so long failed to explain why the West rules is that protagonists have defined their terms in different ways and focused on different parts of the problem. The simple act of setting up an index should therefore move the debate forward. Critics of this book who raise the first of the objections I listed at the start of this appendix—that quantitative comparisons are unacceptable because they dehumanize us—will be forced either to find another way to explain why the West rules or to show why we should not be asking that question at all, while critics who raise objections 2 through 4—that I have defined social development badly, used the wrong traits, or misunderstood the evidence—will be forced to come up with better indices of their own. And then, perhaps, we will see some real progress.
Figure A.3. Even greater error: (a) raises all Western scores and reduces all Eastern scores by 20 percent, and (b) raises all Eastern scores and reduces all Western scores by 20 percent