The same reasoning which applies to uses of the same commodity will evidently apply to any two uses, and hence to all uses simultaneously, so that we obtain a series of equations less numerous by a unit than the number of ways of using the commodity. The general result is that commodity, if consumed by a perfectly wise being, must be consumed with a maximum production of utility.

99

We should often find these equations to fail. Even when x is equal to 90% of the stock, its degree of utility might still exceed the utility attaching to the remaining part in either of the other uses. This would mean that it was preferable to give the whole commodity to the first use. Such a case might perhaps be said to be not the exception but the rule; for whenever a commodity is capable of only one use, the circumstance is theoretically represented by saying that the final degree of utility in this employment always exceeds that in any other employment.

Under peculiar circumstances great changes may take place in the consumption of a commodity. In a time of scarcity the utility of barley as food might rise so high as to exceed altogether its utility, even as regards the smallest quantity, in producing alcoholic liquors; its consumption in the latter way would then cease. In a besieged town the employment of articles becomes revolutionized. Things of great utility in other respects are ruthlessly applied to strange purposes. In Paris a vast stock of horses was eaten, not so much because they were useless in other ways, as because they were needed more strongly as food. A certain stock of horses had, indeed, to be retained as a necessary aid to locomotion, so that the equation of the degrees of utility never wholly failed.

CHAPTER IX

THE LAW OF POPULATION

1. The Movement of Population 1

I. Marriage

Marriage affects the movement of population only indirectly and in so far as, under a system of monogamy, it actually affects the number of births.

It is governed by motives that affect all men, and by others that affect particular individuals; it depends upon external conditions and subjective inclination, upon the economic situation, upon public legislation, and upon ecclesiastical regulation. Like migration, therefore, it is subject to great variation from time to time, and, more than any other phenomenon known to statistics, defies the so-called law of the numerical regularity of social phenomena. In Württemberg, for instance, the number of marriages in the year 1854 was 7905, while in 1871 it was 20,703, or nearly three times as large.

The decrease or increase of marriages sometimes affects economic conditions advantageously and sometimes disadvantageously. An increase indicates that at the time confidence in the future prevails; but in individual cases this confidence may just as easily prove mistaken as justified.

The annual number of marriages depends upon the composition of the population in respect of sex and age. In the long run not more than the regular yearly quota of young men of marriageable age can contract first marriages. If we assume that men normally marry between the ages of twenty-five and thirty, then each year one fifth of the men in this age group

1 By Gustav von Rümelin. Translated from Schönberg's Handbuch der Politischen Oekonomie, I.

will marry for the first time. The number of such men in the German Empire was 7.38 per 1000 inhabitants in 1875, 7.12 per 1000 inhabitants in 1880, and 7.40 per 1000 in 1885, an average of 7.3 per 1000. And in other countries, such as England and France,1 the number does not vary much from these figures, since, as statistics show, the number of persons between the ages of twenty and thirty years is about the same whether population is increasing rapidly or slowly.

But now since 13.6 per cent of all married men contract second or third marriages, it would seem that for Germany the highest annual marriage rate would be 8.3 per 1000 inhabitants. But in reality such a figure could not be permanently maintained, since some men never marry. In the eight years from 1872 to 1879, inclusive, the average number of marriages did rise to 8.80 per 1000, but this was due to an extraordinary number of marriages contracted before the age of twenty-five or after the age of thirty. In other countries during the same period the figures were considerably lower: Switzerland, 7.6; Great Britain and Ireland, 7.3; Belgium, 7.3; Norway, 7; Sweden, 6.6. In France, however, the figures rose to 8 per 1000; but here, as in Germany, the number was abnormally large in the period following the war of 1870.2 During the eighties the marriage rate in Europe was lower and more uniform than in the seventies. In France from 1880 to 1884 the average rate was 7.5 per 1000; in Germany at the same period it was 7.6; in England and Wales from 1879 to 1884 it was 7.4; in Italy for the same years it was 7.6; in Austria the rate was 7.8 from 1880 to 1884, and at the same time in

1 In the United States in 1900 the number of males between the ages of twenty-five and twenty-nine, inclusive, was about 44 per 1000 of the whole population. This would make the yearly quota of marriageable males 8.8 per 1000.- ED.

2 In Massachusetts the number of persons marrying was materially affected by the Civil War :

1860

20.15 per 1000

1863

17.36 per 1000

By dividing these figures by 2, marriage rates can be computed comparable with those in the text. - ED.

Belgium and Switzerland the figures were, respectively, 7 and 6.8; finally in Sweden a rate of 6.3 was maintained from 1880 to 1883.1 It would seem, therefore, that a rate of 8 or more per 1000 can be maintained permanently only in countries where a considerable number of men marry before the age of twenty-five, such as Russia, Servia, and Hungary, or where marriages are easily and frequently terminated, and where consequently remarriages are numerous, as in Transylvania. Great variations above the normal marriage rate are always due to extraordinary . conditions, and are followed naturally by reactionary movements. And just as regularly a marked reduction of the rate below the normal level is a sign of unfavorable economic conditions, and, at the best, a necessary remedy for the conditions produced by the previous excess in the number of marriages.

The average duration of a marriage could be learned, if at all, only from family records, since it cannot be ascertained by a census. . . . It can be estimated, but only approximately, by gathering data for a series of normal years and then dividing the number of married couples by the number of marriages dissolved by death. The duration will, of course, be longer for early marriages than for marriages contracted later in life. According to the estimates of Wappäus, the duration of marriage is from twenty-one to twenty-six years; and for Germany and the countries of Middle Europe may approximate twenty-five years. The same thing holds true of the average duration of fertility in marriages, that is, the average difference between the age of

1 In 1900 the marriage rates in various European countries were as follows:

In the New England States for the period 1893-1897 the marriage rates were:

These rates are above the average for Europe, probably because "the sexes are quite equally distributed and the proportion in the middle-age groups is large" (Bailey, Modern Social Conditions, p. 137). — ED.

the first child and that of the last child born of a marriage. Direct information could be secured only from family records. From such a source it has been learned in Württemberg that the period of fertility in the average marriage approximates twelve years, and a computation based upon data taken from the Almanach de Gotha tends to confirm this conclusion.

Associated with these statistical concepts is that of a generation, by which is meant, not the total number of persons living at any one time, but the time required for one generation to succeed another, the average difference of age between parents and children. The length of a generation, then, is computed by adding to the average age at which men marry one half of the average duration of fertility. For countries of early or later marriages, as well as of larger or smaller birth rates, it ranges between the somewhat wide limits of thirty-two to thirty-nine years. For the countries of Middle Europe it averages from thirty-four to thirty-five years. The length of a generation influences materially the social and political development of countries, since the rate of change is more rapid if the life of a generation is short, and is retarded if the life is longer.

From statistics based upon several million marriages in countries of Middle Europe it has been learned that, upon an average, 811 out of every 1000 marriages were first marriages for both of the contracting parties; that 106 were marriages of widowers to maidens; that 53 were between bachelors and widows; and that 30 were between widowers and widows. In this computation, however, divorced persons are counted as widowers or widows. In periods of economic depression it appeared that the proportion of marriages of widowers and widows increased because such persons, as a rule, already have an assured livelihood and do not have before them the struggle to establish themselves in the world. As an example of this tendency it suffices to state that in Württemberg between the years 1838 and 1857, a period which included the years of depression from 1845 to 1855, out of every 1000 marriages 185 were those of widowers and 81 were those of widows; whereas from 1871 to 1880 the marriages of widowers averaged 146 and those of widows, 67.