Godwin, William. Of Population. London: Longman, Hurst, Rees, Orme and Brown, Paternoster Row, 1820.

CHAPTER VI.

OBSERVATIONS ON THE SWEDISH TABLES CONTINUED

BUT there is another view of the subject, equally worthy of notice, and well calculated to throw light upon the topic before us.

I have just stated that the annual number of marriages in any country, cannot, for any length of time, exceed the number of females annually arriving at a marriageable age.

Now let us take this question in another way. Thought I have set out with considering the women capable of child-bearing as the soil or nidus in which the successive generations of mankind are reared, yet it is equally true, that husbands are necessary to the consummation of marriage, as that wives are so, and, at least in countries where polygamy is forbidden, that there can be no more marriages than husbands.

The same inference therefore should seem to follow as to males, which I have already drawn as to females, viz., that the annual number of marriages in any country, cannot, for any length of time, exceed the number of males arriving at the age which it is permitted, or rather at which it is usual for them to marry

[177]

But the number of males, though they are born in greater numbers, will be found at almost any age above childhood in all Tables of Population, and specially in those of Sweden, to fall short of the number of females.

In Sweden, the country we are here considering, there is a law, forbidding any individual of the male sex to marry, till he has completed the twenty-first year of his age. ^a

To this consideration it may be added, that it will scarcely happen, that every male will be disposed to marry, as soon as he has completed the twenty-first year of his age. Perhaps, reasoning on this principle, the marriages which annually take place in Sweden cannot, for any length of time, be expected to exceed the number of males who annually arrive at twenty-five years of age. This will reduce the number of marriages, and consequently increase the number of females who spend their lives in the single state.

Such would appear at first sigh to be the speculative principle of the subject, and would contradict what has been established respecting it in the former chapter. But let us see how it stands, as practically exhibited to us in the Swedish Tables. And here as in the case of the fe-

[178]

males, I will take the fifth par of the males between twenty and twenty-five in the year under consideration, as the number arriving in that year at twenty or twenty-five years of age. The number arriving at twenty-five will indeed be less than the number arriving at twenty, in proportion to the males who are found to die annually between those periods of life. But this is not the season of human existence most considerably exposed to the accidents of mortality; and I will wave for the present the taking that diminution into the estimates.

The three years then, 1757, 1760, 1763, as appears from the Tables, will stand as follows:

It has already been observed, that the females becoming marriageable do in the most years exceed, as we should expect them to do, the annual number of marriages. For certainly, the marriages of any one year do not form a standard: the marriages of any one year may exceed: my proposition is, that the annual number of marriages cannot, for any length of time, exceed the number of females annually arriving at the marriageable age.

[179]

Add to which, I have taken the marriageable age at twenty; but it is possible to marry before that age; and the Swedish law permits females to marry at fifteen. _b Now the number of females annually arriving at fifteen is greater than the number of females annually arriving at twenty. If therefore the number of marriages exceeded the number of females annually arriving at twenty, the excess must necessarily be supplied form the females between fifteen and twenty.

But the case of the males is different; and they, as I have said, are forbidden to marry till they have completed the twenty-first year of their age. How then are we to account for the excess of marriages above the number of males annually arriving at twenty-one?

This difficulty will be found to be in a considerable degree removed by an inspection of the Upsal Table. _c Few things are more striking in this Table than the excess of the number of widows above that of widowers. Adding together the whole series of nine years there exhibited, the number is

of widowers 20,567

of widows 108,537:

the number of widows being more than five times the number of widowers. But married women, as may be judged from the Tables of Sweden in general, die with nearly as much

[180]

rapidity as married men. The small number of widowers can therefore only be accounted for, by the infallible inference, that five times as great a number of widowers as of widows, are found to marry again. And form the same principle we are entitled to conclude, that they intermarry generally, not with widows, but with virgins, or what our law calls spinsters.

To apply this, let us observe that, if the diocese of Upsal in 1793 contained 11,874 widows, and the whole of Sweden by the rule of proportion would appear to have contained 135,712. But if we suppose as many men to have lost their wives as women to have lost their husbands, it would then follow that upwards of 108,000 men had married a second time, even without taking into account those who might a second time have become widowers. This affords an ample allowance for the deficiency there might otherwise appear in the number of marriageable males.

Having referred in this place to the Table of Population for the Diocese of Upsal, I will here comment upon one or two particulars in it, which seem to require explanation. This Table descends to a greater fullness of distinction and enumeration than any other that has fallen under my observation; and it is therefore particularly desirable that it should be well understood.

One circumstance which appeared to me at the first view somewhat surprising, was the small number of households in the last column, compared with that of the subsisting marriages in the fifth. This indeed is in no way material to the question I am investigating; but it is right for the satisfaction of the reader that it should be cleared up.

I stated this difficulty to the intelligent Swede,^d who had the goodness to assist me in translating the heads of these Tables; and his explanation was as follows. "By a household or establishment we understand all those persons who eat at one table, or, more properly who are subsisted from one income or expenditure. For example, at Sir Joseph Banks's there are various tables at which different persons are fed, but the whole expense is defrayed by one individual. This therefore is one household. If, on the contrary, there are several families dwelling under one roof, but which are, so to express myself, not nourished from one common root, these would be counted in the Swedish enumeration as separate households. Now in this country ['Sweden'], nothing is more common, particularly in the rural parts, than for the sons, after they are married, to live under the roof with their father, all together constituting one ample household. This is the reason why, in the Table of Population for the Diocese of Upsal, there appears so much smaller a number of households than of subsisting marriages."

Another circumstance which may need elucidation is, that the number of unmarried males and females above fifteen years of age, in the eighth and ninth columns, may appear at first sight greater, than from previous reasonings might have been expected.

Upon this I would remark, first, that it is not rational to suppose that there can be any substantial discordance between the Tables of Population for Sweden generally, and the Tables of Population for one of its most considerable provinces. The comparisons I have exhibited between the number of annual marriages and the number of females annually arriving at twenty, are expressly taken out of the Tables of Population for the kingdom of Sweden.

Secondly, every reader will perceive that there is a vast difference between the setting down in figures on the one hand, the number of females arriving at twenty in any given year who shall finally remain unmarried, and on the other the setting down the number of females at all ages, who at any given period shall be found unmarried, though then may happen to marry in the next year or the next week. The number in the last case may be great, at the same time that number in the former may be exceedingly small.

Thirdly, the unmarried in the Uspal Table include all who have passed their fifteenth birthday at which age according to the Swedish law females are permitted to marry. But in the extracts I have made from the Tables of Sweden in general, I have taken the marriageable age at twenty. Therefore the Uspal Table swells the number of the unmarried females by the whole amount of those between fifteen and twenty, or at least by the amount of such as shall not have married between those periods. But the females between fifteen and twenty will be found to constitute nearly a twelfth part of the entire female population.

The proposition which I deduced from the Tables of Sweden in general, is that the annual marriages nearly equal in number the females annually arriving at twenty; or, in other words, that there are nearly as many women married every year, as there are women arriving every year at that age. The only limit upon that proposition would be in the number of women who shall end their lives in the unmarried state.

But the column of unmarried females in the Uspal Table, does not set before us the number of females that shall live and die unmarried. In the first place, it may well be supposed that the greater part of the females between fifteen and twenty, making a twelfth part of the entire female population, will hereafter marry. In the second place it is to be considered, that the total amount of unmarried females in any kingdom or province at a given period, will materially depend upon the customary age of marriage. If every female throughout the state married the day she completed her fifteenth year, then it is self-evident that the column of unmarried females above fifteen would be left a complete blank. But, if on the other hand the marrying age were from fifteen to thirty-five, then all might marry, and yet half the females between fifteen and thirty-five would constantly appear in the column of the unmarried.

Another consideration is to be added, which I may thus illustrate. Let us suppose the females annually arriving at twenty to be 20,000, an that of these 19,000 marry, and 100 continue in the single state. Let us suppose that there is some natural reason, of infirmity or otherwise, why this twentieth part of the female division of the community should not marry. There would thus be 1000 females to be placed in the column of the unmarried, for the year for which this account is taken. In the next year there would be one twentieth of the females arriving at twenty in that year, or 1000 more, to be added to the 1000 of the preceding year, except so far as this last number was diminished by death, and so on ad infinitium. Thus, as we said before, if every female throughout the state married the day she arrived at the marriageable age, the column of the unmarried would be blank; but, if one twentieth remained unmarried, and continued so, this in time would amount to one twentieth of all females living in the state, who were beyond the marriageable age. It is unnecessary to say more on this point: every reader who is desirous of so doing, will be able to follow out the further particulars for himself.

There is another circumstance entitled to our consideration, before we finally determine what degree of authority is to be attributed to the Swedish Tables. In the reasonings I have exhibited, I have set down the women capable of child-bearing as one fifth of the whole community. At the same time it fully appears from the Tables, that the births of scarcely more than four to a marriage. Now, if of the number of the born only one in five is to be counted on to become a mother and give children to the next generation, it clearly follows that the number of women capable of child-bearing will in each successive generation perpetually diminish, and consequently that a population so circumstanced must be regularly advancing towards utter destruction. But the Swedish Tables, from which these two facts are taken, exhibit a progressive increase of the number of inhabitants. Either therefore this apparent contradiction must be reconciled; or the Swedish Tables must be admitted to be an imperfect authority on which to rest our conclusions.

In answer to this difficulty I would observe, in the first place, that one of the most irresistible results of the Swedish Tables, is that there are four births to a marriage. But this proposition, if true, must be equally true if taken in an inverse form, and we state it-to every four births there is a marriage, or, in other words, for every four births there is a marriageable woman. One of these propositions cannot be true, and the other false; and the number of women of an age capable of child-bearing is hereby, clearly established.

Secondly, it is proper to observe that, though it was sufficiently reasonable to set down, as the foundation of our inferences, the period in which a woman is to be considered as capable of child-bearing, as beginning when she is twenty years of age, yet this proposition is by no means absolute and uncontrollable. The Swedish law admits of the female marrying at fifteen; and as, necessarily, more human creatures live to attain the age of fifteen than of twenty, we have here a considerable addition to the stock of possible mothers. The females between fifteen and twenty form a sort of corps de reserve, from which the brigade of marriageable women may be recruited in case of necessity.

Thirdly, it is to be remembered that we found the number of births to a marriage exceeding the amount of four by a small fraction.^e Now this fraction may be first sight appear scarcely worthy of notice, yet, in its operation over a nation consisting of three millions of souls, and spread over a succession of year, it would doubtless have the effect of rendering that population progressive, which without this fraction would have been stationary. There is therefore nothing contradictory and irreconcilable between the different particulars exhibited in the Swedish Tables.

Here then we are presented, as far as it goes, with a solid basis of reasoning concerning the possible increase of the numbers of mankind. Of every other country in the world we may be said in this respect to know nothing. In Sweden great labour has been continued through a series of year; and it has been prosecuted on the most enlightened principles. We learn therefore from this example, perhaps as nearly as possible, how fast the race of mankind, at least as society is at present constituted, can increase, beyond what limits of pace and speed of multiplication cannot be carried.

Sweden is a country in every respect as favourable to the experiment as we could desire. Almost all the women marry. "The continual cry of the government," as Mr. Malthus expresses it, "is for the increase of its subjects.^f" And the soil is so thinly peopled, that it would require many ages of the most favourable complexion, for the inhabitants to become so multiplied by the mere power of procreation, as to enable them to rear and to consume all the means of subsistence which the land might easily be made to produce.

^a It is however allowed to a person of the male sex to marry at eighteen, provided he has any landed property, holds any office, or has in any other way the visible source of a regular income. Handbook i Svenska Kyrkelagfaranheten, or Manual of the Swedish Ecclesiastical Law, Chap. 1 § 6.

^b Handbok, ubi supra.

^c Table VI, p. 158.

^d Mr. Nairman, one of the librarians to Sir Joseph Banks.

^e Page 172.

^f Vol. I, p.391.