STATISTICS OF POPULATION 1. Demography may be defined as the application of statis tical methods to the study of populations. As such its scope is as broad as the activities of mankind. Our interest in it is exceedingly narrow, only relating to those statistics which indicate the changes and fluctuations of a population, as well as the degree of its health. This narrow field of demography is commonly designated as vital statistics. The subject of statistics may be defined as the numerical statement of facts for study and comparison, and the collection of data and its abstraction for generalization.
2. All studies in vital statistics are based upon the population of the area under consideration, and are necessarily expressed in terms of population. Statistics of population for the pur poses of vital statistics must show the number of inhabitants of an area, classified by age, sex, nativity, race and occupation. More detailed information would be desirable but the labor of making more detailed analyses is commonly too great for the available funds.
This fundamental information regarding a population is obtained by a census enumeration. The standard census in the United States is conducted by the Federal Government. These have been taken every TO years since 179o. In addition many states take decennial censuses at intervals mid-way between the years of the Federal census. In order to avoid omissions and errors it is desirable to take the censuses at a time of the year when the maximum number of individuals will be at their homes. The spring is considered the best time of the year for this purpose. Actually the census canvas requires several months for its accomplishment, yet for simplicity sake it can be, and is referred to a given date. Thus the 1910 census was referred to April 15th.
Of errors in census enumerations, the most important are the following: (1) Overlooked persons. It is not possible to require that the entire activities of a nation be suspended while the great inventory is being taken. Some individuals, espe cially those who are traveling are bound to be overlooked, others may be counted twice and unscrupulous enumerators or officials in charge of areas where a great civic pride in rapid population increases exists, all contribute their quota of error. On the
whole however, the Federal enumerations probably represent the most accurate enumeration possible, and the actual error is but a small fraction of one per cent. These censuses should always be taken as the standard. (2) A second common error results from misstatements regarding age. These are of two types. The 'age recorded should be the age of the individual at the last birthday. In the case of children up to five it is very common to find the stated age is really the age at the next birthday. On the other hand adults show a tendency to express their age in round numbers, thus a person of 59 will give 6o as their age, or 6o for 62. As a consequence when the population is plotted according to age we find a great increase in the num ber of persons at each of these decennial ages over and above those at intermediate ages (Fig. r i9).
Populations are never stable, continuous fluctuations is the rule either increasing or diminishing (Fig. 123). The most important as well as the most variable factor in these fluctua tions are migrations,. either a moving out (emigration) or a moving into an area (immigration) (Fig. i2o). Migratory populations show a large relative proportion of adult males. As a result of these changes we may find that the entire charac ter of a population has changed, or its distribution within an area has been altered. Immigration is an important factor in territory under-going industrial development, while emigration is more commonly observed from areas of industrial stagnation or where competition is keen and sharp.
The other constant factors in population fluctuations are the increases resulting from births and the losses arising from deaths. Where the effect of migrations are not felt, a gradual increase is usually observable, the increments from births being usually in excess of the death losses (Fig. 121). On the other hand, in some countries, such as France, the death rate may nearly coincide with, or at times exceed the birth rate (Fig. 122).