THE PRECEDING CHAPTERS have demonstrated the appreciable rise in the level of living made possible by technical progress in France during the past two centuries. We turn now to the situation of the other countries of the world. Does France occupy a privileged situation? Has there been a like improvement in the purchasing power of wages in other countries, and if so, to what extent?
The answers to these simple questions would have been difficult to supply only 15 or 20 years ago. The credit for pointing out the scientific and human interest of comparative studies of the level of living belongs to the economic staff of the League of Nations. However, it was Colin Clark who had the honor to first prepare a general table of the levels of living achieved in the modern world and to state the essential questions about the causes and conditions of the observed discrepancies (in The Conditions of Economic Progress, 1940). Even if he did not distinguish all of the conditions and even though he may have been mistaken about some of the causes, his book will always have an important place in the history of economic science. In proposing other formulations than those of Mr. Clark, and even in criticizing many of his principles, we are only exploring a territory that he pioneered.
Since 1940, by the way, the available statistics about level of living, national income, and purchasing power have become much more abundant. These measures permit a ready answer to the question raised at the beginning of this chapter about the relative situation of France among the nations. Many nations today have a level of living far above that of France; many have a level of living far below.
Some of these, like India and China, have a level of living not much different from France in 1800. Others have a level of living two or three times higher than France in 1950. All of this comes about as if the nations were spaced out along the route of time. The development that France has undergone since 1750 is one pursued by all of the nations of the world, but some of them mount the ascending slope with great speed, and others very slowly. Several nations are in advance of France, the majority remain behind. Thus, the time differences in the speed of the process explain the spatial distribution at any given moment.
These facts will appear clearly in the following four sections which describe successively the four fundamental aspects of the problem of comparative levels of living: nutrition, average total income per capita, real earnings, and indexes of consumption.
Half of today's world is still living in the age of millet. Two-thirds of the human beings now alive are undernourished. These incontestable facts emerge from all of the contemporary studies and especially from the simultaneous investigations of several agencies of the United Nations.
The normal average ration of a normal population consisting of children, adults and old people is, as noted above, from 2,700 to 2,800 calories per head. The total agricultural production plus the possible importations furnish something less than 2,250 calories per head in enormous regions that include more than half of the earth's surface. These are almost all of Asia, a large part of Africa, some parts of Central America and South America. Only three-tenths of all humanity have more than 2,750 calories daily. These are the great nations of the West, the Soviet Union, and the British dominions. The general situation may then be summarized in the following table:
The U. N. experts have estimated the average nutrition of those nations which were willing to furnish basic statistics as shown in the following table. (The original numbers have been rounded.)
If we remember that 2,000 calories corresponds to about 850 grams of wheat per capita and per day, it is evident that the two most numerous populations of the earth, the Chinese and the Indians, occupy a situation with respect to nutrition which is very similar to that of the French under Louis XIV.
The differences in level of living in the modern world are plainly considerable. The average differences are greater in some cases than the differences among individuals in the same nation. There are not 200,000 Frenchmen as badly nourished today as the average of 700 million Asians. The daily food of an average Chinese or Indian would cost 50 francs in France, which is the wage paid for 20 2minutes of a day laborer's work. Put another way, every French laborer would be able to provide himself with the average ration of an Indian by working only 120 hours a year. Inversely, if a Frenchman were to offer 120 hours of gratuitous work to a Chinese, he would be able to double the latter's ration; if this same Frenchman were willing to limit himself to the same diet, he would be able to accept the responsibility, in working a normal year, for no less than 19 Chinese, so that, making all due allowance for costs of transportation, the 21 million French workers would be able to double the food allowance of the 400 million Chinese by working 19 out of 20 hours for the Chinese.
This example is intended only to illustrate the extraordinary differences that have developed between the levels of living of different peoples. The study of real per capita income confirms the existence of these disparities.
The best estimate of the average level of living of a population is obviously that which is based on an estimate of the real national income per capita.
Although the measurement of nominal national income, expressed in current money, has made great progress since 1940, this measurement still includes, as may be imagined, a large component of error. Moreover, the conversions necessary to translate nominal national income into real national income are themselves productive of sizable errors. The facts adduced in the previous chapters show that the structure of consumption varies with the volume of that same consumption, and that the prices of different goods are very far from any parallel evolution in time or any stable proportionality in space. It is quite impossible to develop any unit of consumption that is valid both in time and in space. Furthermore, it is impossible in principle to establish from year to year and for each nation the price of a standard consumption unit that will remain comparable between countries and from one era to another.
I have discussed in another book the important problem of real income. Here it is sufficient to recall that the concept of real income has a tendency to dissolve as we attempt to make it more exact. This does not mean, however, that real incomes cannot be compared when the differences are of sufficient magnitude. It is possible to chart large scale changes in real income by relating nominal income to the cost of a fixed cluster of currently consumed goods, selected to be as representative as possible of the average habits of a given people at a given time.
We shall examine first the orders of magnitude of the estimates given by Colin Clark for the past fifty years. We shall then present additional evidence about the recent development of the level of living in different nations, based in part upon the studies of the United Nations, and in part upon the investigations of such contemporary economists as Kuznets.
Colin Clark has expressed the nominal national income per capita of a very large number of countries in international units of "constant purchasing power," by estimating the price of a defined cluster of goods and services in each country and at each epoch. He arrives thus at a table of equivalence of currencies like the following:
One U. S. dollar, 1934, equals 15 French francs, 1934; or 17 French francs, 1935; or 30 French francs, 1936.
Working in the opposite direction from such a table, he is able to provide an estimate of the real national income for all of the countries of the world and for all the years considered, expressed in 1934 dollars. As a matter of fact, the unit of reference that he uses is the average purchasing power of an American dollar in the period 1925 to 1934. Despite the serious reservations that must be made about this method of converting, using as a base a cluster of commodities that may correspond to the real habits of one population and not at all to the habits of another population or another area, it is nevertheless certain that the calculations made in this way are useful as first approximations, the more so because Clark's method is relatively unfavorable to the rich countries in comparison to the poor countries. In poor countries the gap between the cost of luxuries like meat, and the cost of staple products like cereals, is much less than in rich countries. For example, in France in 1850, an average pound of meat did not cost more than three times the price of a pound of bread, while in modern France meat is worth at least six to eight times its weight in bread. There is nothing complicated, of course, about this phenomenon. It does not make sense to say that one people has a level of living which is twice that of some other, since in reality no people consume twice as much grain, twice as much milk, twice as many vegetables, twice as much meat at the same time. Nevertheless, the figures in Table XVIII provide a rough but suggestive idea of the distribution.
The figures in this table confirm the scale of the international differences previously noted with regard to the level of living. Between China and the United States, the difference is of the order of 1 to 12. It would be 1 to 30 if it were measured with reference to rice or wheat. Still more impressive, if purchasing power is measured with reference to a manufactured product such as a bicycle, a radio set, an automobile, or an electric motor, the disparity between China and the United States will increase to 1 to 50, 1 to 80, or even 1 to 100. On the other hand, with regard to such commodities as haircuts, legal fees, medical expenses, or handmade objects, the differences will be much slighter and sometimes even to the advantage of the Chinese, although these are precisely the goods and services that the Chinese is hardly able to obtain because all of his income is devoted to food. It little matters that some services are cheap if, in fact, they are out of reach of the mass of the people.
These remarks show at the same time how much in these figures is arbitrary and how much is representative. In scientific terms, they can only be summarized by some such statement as the following: the statistical studies of real national income confirm the studies of nutrition alone. The differences existing at the present day between the nations are such that in some countries, most people do not have enough to eat because all of their income is devoted to the consumption of cereal, while in other countries, the average man not only consumes more expensive foods, but also a host of other products and services, so that the cereals which take more than 80 per cent of the budget of the average Chinese do not absorb more than four per cent of the budget of the average Frenchman and two per cent of the budget of an American.
2) Short term changes in real national income
The preceding facts, it seems to me, do not leave the slightest doubt about the fundamental direction of contemporary economic evolution: a long term rise in everybody's level of living. Some nations have barely begun this development, due primarily to their huge population growth, but in all the countries of Christian civilization, it was well under way before 1900. The speed of improvement may vary considerably. The general direction is the same everywhere.
However, it is obvious that the movement has little or no regularity. It is often the case that the short term shows regression in relation to the long-term secular trend. Some of these regressions are due to economic crises such as unemployment. Others, much more serious, are due to war or revolution.
The following table shows the regression due to the crash of 1929, and the regression in Germany and most of the other countries of Europe, caused by World War II.
Tobie XIX-Recent Trends in National Income and the Cost of Living in Several Countries
The figures show that the regressions attributable to the crisis of 1929 reached an order of magnitude of 25 per cent.
Those due to the war were very much greater. Hungary experienced, in 1945, a fall of 50 per cent from the 1939 level. The findings about real national income per capita are further confirmed by studies of real earnings in various countries.
3. REAL EARNINGS AND THE PURCHASING POWER OF WAGE EARNERS
The study of real earnings is, as we have seen, complex. To the difficulties arising from the duration of work, the skill of the workman, and the nature of the goods consumed, there must be added for international comparisons the difficulties that arise from occupational organization, methods of work, and customary wage arrangements, involving such matters as perquisites, the "thirteenth month," family allocations, social welfare payments, taxes and so forth, as well as differences in the traditional style of life.
We shall not return here to the principles which we developed in the Civilization of I9754 concerning the fundamental distinctions between hourly wage rate, daily wage rate, and annual wages. We shall not return, either, to the comparisons based on the prices of a quintal of wheat, which show great disparities from country to country.
Instead, we shall give other evidence on real earnings which is less familiar and less accessible, and which confirms our previous conclusions.
Since 1926, the Revue Internationale du Travailhas been publishing a very remarkable series of statistics on the wages observed in twelve of the world's great cities, and on the prices in each of these cities of the "basket of food" of working families. As baskets of food are not composed of the same assortment in different countries, the study has used five types of baskets. Francois Simiand prepared a paper on this investigation which, as far as we know, was never published. The accompanying graph is taken from that paper and is based upon figures furnished by the Revue.
Between Milan and Philadelphia, the discrepancy was of the order of l to 3.5. The basket in Great Britain (the basket of a relatively rich worker) costs relatively more than those of other and poorer countries and less than those of richer countries. The difference in the price of different baskets is not very great except in Australia where it reaches 30 per cent.
At a more recent date, Lehoulier has published in the Bulletin of the General Statistical Office of France an excellentstudy on real weekly wages in various countries.7 Lehoulier has found the average of weekly earnings in the various nation for six well-defined occupations in six industries-coal mining metallurgy, construction, lumber, textiles, and printing. H related these nominal salaries to the price of a certain number of food staples taken in definite quantities.
The results obtained by Lehoulier are summarized in Table XXII, which compares the purchasing power of each group to that of the French workman of 1914 taken as 100.
The difference between the United States and the other countries is immediately evident. England and Sweden form another group still very favored in relation to the third, consisting of France, Germany, and Italy. We see further that, while the three first ranking nations gain from 2-0 to 30 per cent between 1914 and 1938, the three others have either gained very little, like the 15 per cent for France, or even lost more than 10 per cent for Italy. Finally, it should be noted that the great progress of the United States was accompanied by a marked sensitivity to crises. From 1929 to 1932, the purchasing power of the American worker fell from an index of 271 to an index of 182, or 30 per cent. This regression is greater than that shown by the real national income. It occurs because Lehoulier's index uses industries that are especially vulnerable to crises. An index of earnings of white collar workers in commercial establishments would have shown somewhat less decline. Whatever the exact amount of this regression may have been, it seems certain that severe short-term regressions were associated with strong long-term progress. The most progressive economies appear to be the most fragile. However, it is worth remembering that even at the worst of the depression of 1932, the index of purchasing power of the American worker stood at 182 compared to the maximum of 120 which the French worker achieved in 1935.
Dessirier has published a series of studies on the real earnings of various nations. He estimates at 10 to 20 per cent the possible error of his figures. These calculations confirm the findings of Lehoulier, and furthermore they apply to a number of other nations. Using France, 1913, as the base of 100, the indexes reached the following orders of magnitude in 1939 and 1944:
It must be noted that all the earnings entered in the above calculations were full-time earnings, without any allowance for partial or total unemployment. The national income per capita, on the other hand, automatically takes account of unemployment in the nation.
All of the foregoing calculations, those of Simiand, Clark, Lehoulier and Dessirier, are based alike on expenditures for food. We have sufficiently stressed the point so that the reader will certainly recall that one arrives at entirely different figures in measuring purchasing power against manufactured objects, clothing, rentals, services, and so forth.
On the other hand, it is certain that the consumption of food, in "baskets of provisions," gives a vivid and socially relevant image of the facts.
The correlation of the average real income per capita and of average real earnings must be explored. This has been done by Brousse. The findings are fairly consistent and permit the following comparisons for the period 1925 to 1934:
This table serves to confirm both the disparities that exist between countries, and the fact that the purchasing power of wages determines-at least approximately, and at least in industrialized countries-the average level of living of the nation.
The preceding estimates may be further checked against numerous indexes of apparent consumption.
These various indexes confirm, on the whole, the facts already placed in evidence in this chapter. Hardly a third of the human species now alive can be said to have a level of living that surpasses the simple satisfaction of minimum food needs. The other two-thirds are still in the vegetative phase of history.
France belongs to the first third, but in spite of the progress made in the last 200 years, she is far from the head of the movement. Among the nations comprising this first third, and which are considerably dispersed from the United States at one extreme to the Soviet Union at the other, it can hardly be said that France occupies an average position, for, in fact, she seems to be closer to the Soviet Union than to the United States.
The most striking result of this inquiry is that in 1956, nations still existed whose level of living had remained practically the same as during the long centuries of the traditional era, while at the same time others had transformed their living conditions. The world of today simultaneously contains communities that have hardly evolved at all in 2,000 years and others that are in full transition to a new human condition.
What are the causes of this progress? Why have certain peoples been able to reach and maintain a pioneer position? Why are the others barely capable or even incapable of imitating them? What is the dynamic element in contemporary economic evolution?
It is customary to give arbitrary indexes for the level of living of populations. However arbitrary, they convey a clear picture. We know by various means-through tax records or direct enumeration-a whole series of material facts that are unquestionably related to the level of living and the purchasing power of nations.
Table XXIII gives certain statistical fragments of this kind. The figures invite commentary and criticism. Some of them are representative and exact. Others are questionable. Such as they are, they nevertheless bring out orders of magnitude that cannot be challenged.
The four first columns of the table present estimates of the consumption of food items typical of rich nations: sugar, tobacco, tea, coffee, cocoa, oranges, tangerines, bananas. The longitude of the country involved must, of course, be taken into account. The consumption of sugar is one of the most informative items. We saw above that in 1880 the French consumed 8.6 kilos per capita per year, the same as the present day consumption in Italy. The United States, Great Britain, and the British Dominions were close to 50 kilos per capita per year in 1938, about double the figure for France. During the occupation and postwar rationing, the per capita sugar allotment in France was 6 kilos per year.
. See also, C. Clark, The Conditions of Economic Progress; Office of the United Nations, Enquête mondiale sur l'alimentation (1946); Programmes européens de remise en état et d'amélioration de l’agriculture (1948); International Emergency Food Council, Reports of the Secretary General to the Meetings of the Council.
 For a discussion of the errors inherent in estimates of real national income, see J. Fourastie, "Sur la mesure des quantités économiques", Revue d'Economie Politique, January, 1956.
 J. Fourastie, Productivity, Prices and Wages (1957) .
 La Civilisation de I975, p. 44.
 July, 1926.
 We wish to thank Georges Luftalla, to whom Simiand gave his files, who permitted us to use the unedited writings of this great economist, and who authorized the publication of this graph on the level of living.
. Bulletin of the General Statistical Office of France, June-September, 1944.
 The Bulletin of the U.S. Bureau of Labor Statistics published early in 1950 an evaluation of the purchasing power of one hour's wages for food, with the United States equal to 100. The range is from 109 for Australia to 18 for the U.S.S.R. France has an index of 35. China and India are not in the table.
The study also indicates "how much time a workman must work to receive a certain product.'' It can be seen that the spread between the rich and the poor countries is weaker for food than for manufactured products, and weaker for housing and services than for food”.
There have been numerous publications on this subject in the past several years. See especially Margaret J . Hagood, "Exploration of Technics for Measuring Economic Density of Population," Actes du V- congres mondial de la population, Vol. Ill, pp. 315-31.
 H. Brousse, Le niveau de vie en France (Series, Que sais-je?), p. 15.