Footing the columns, obtaining the grand total, and dividing by the whole number of items marked, an arithmetical average is obtained which is similar in form to index numbers in general. May it be accepted as in fact an index number of social pressure for the Commonwealth described?

The most obvious objection is that the enactment of a law does not necessarily mean enforcement. If we had comprehensive and accurate statistics of prosecutions and convictions and of orders of injunction and mandamus, it would of course be desirable to subdivide our categories further, so as to show gradations of enforcement. Statistics of this kind, however, are notoriously defective, and an attempt to mark degrees of enforcement would introduce the uncertainties of subjective estimate.

I am satisfied that the objection itself is not a serious one. The community has other ways of enforcing its will which supplement the legal process. There is an immense amount of coercion which is exercised through public opinion, social favor or disfavor, and the conceding or withholding of business opportunity; and observation will, I think, confirm the judgment that, when the community or some interest in the community feels strongly enough upon any given subject to express its will in statute law, it can and does in a large measure enforce its decree in extra-legal ways, even when the statute itself is enforced but imperfectly. I conclude, then, that the actual enacting of a statute by a legislative body is, in practically every case, an index of actually existing social pressure.

It should be explained, perhaps, that this is true even when the pressure originates with a minority or even with a few individuals, since even then it is an actually disturbing influence affecting in some measure the liberty of all other individuals upon whom it impinges.

The simplest way to obtain such an index number as has been described would be to count up from the statute books all the acts of any category, for example, the category "Prohibited," and falling within a given range as to penalty, multiply this number by the grade number at the head of

the column, enter the product in the column, and so proceed with each subdivision of each category; that is to say, with each column. This method would have, however, the very serious objection that the investigator's procedure could not be followed and checked by his readers, and mistakes of classification and of counting could therefore pass undetected.

The method of full detail (that is, of entering at the left-hand margin of the table every item of legislative enactment) is open to the quite different objections that it is exceedingly cumbersome, and that it leaves much to be desired in respect of possible groupings of similar things, whereby the incidence of social pressure upon various phases of individual or of collective action might advantageously be exhibited.

I prefer, therefore, a scheme according to which a rather full list of items, grouped according to certain sociological categories, is presented in the left-hand margin of the table. Supplementing this list should be the items "All Other Prohibitions," "All Other Restrictions," "All Other Regulations," etc. The number corresponding to each of these items (e.g., the number of "All Other Prohibitions" as counted) should be multiplied by the grade number at the top of the proper column, -penalties having been averaged, and the product should be entered in the column.

The objection may be made to this last procedure, as it might be made to the method of complete detail, that in comparing one commonwealth with another we sometimes find that in one commonwealth prohibitions are itemized which in another commonwealth are brought under one general prohibition. We, therefore, might have three or four entries of any given number in one table, and in another table only one entry, the facts being substantially the same in the two cases, except in respect of one point, statutory detail. I am disposed to think that this complication does not vitiate the procedure, because I believe that the specific itemizing of legal commands is itself usually an index of social pressure; that is to say, it is usually a higher degree of social pressure which makes detailed or specific prohibitions or demands, instead of general ones.

There is one final and important reason for making any such table rather detailed. Having a large number of numerical entries in the columns, we may advantageously obtain all deviations, item by item, from the mode and from our final average or index number, and get the standard deviations. These, showing the degree of closeness of clustering about the average and the mode, are index numbers of some value, inasmuch as it is a generally accepted corollary of the principle of evolution that variation from type diminishes with increasing environmental pressure upon the type. In terms of social causation, this means that increasing social pressure, manifesting itself in law, tends to create uniformity of conduct. Some indication of the extent to which this actually happens should be afforded by the standard deviations.

I hope in a subsequent paper to present a number of detailed tables, with actual index numbers of social pressure, for certain American commonwealths, and in connection with them to touch upon various questions of interpretation.

THE BILL FOR THE THIRTEENTH CENSUS.

BY WALTER F. WILLCOX.

The most important piece of legislation regarding the Federal Census between the insertion in the Constitution of the provision requiring a census and the present time is the law passed in 1899, providing for the Twelfth Census of the United States. The Censuses of 1880 and 1890 had been so overloaded with subsidiary investigations that both the accuracy and the speed of the fundamental census inquiries had been impaired. The law of 1899 divided all inquiries which theretofore had been taken in connection with a decennial census into two classes, those requiring the assistance of the army of Federal enumerators appointed every ten years, and those not requiring such cooperation. The former class included only the census of population, of deaths, of farms, and of manufacturing establishments. The latter class included all other census inquiries. The law of 1899 required the inquiries of the former class to be pushed to completion before any inquiries of the latter class were begun, and set a date-July 1, 1902-at which the inquiries of the first class must be finished. These instructions were carried out by the Census Office, and the main results of the census proper were given to the public with a speed previously unknown. But this legislation incidentally paved the way for the establishment of a permanent Census Office,-a step which was taken in 1902. That office was charged with carrying on the inquiries which at previous censuses had been taken in connection with the decennial count, but which had been postponed under the legislation of 1899 until after the census proper was finished. The permanent Census Office, established in 1902, has been engaged from that time to the present largely in the prosecution of inquiries which have heretofore been taken as a part of a decennial census. Such inquiries include the Report on Wealth, Debt, and Taxation (1907); a continua

tion of the volumes on the same subject included in the Census of 1880 and 1890; the Census of the Defective, Dependent, and Delinquent Classes,—namely, the Blind and Deaf (1906); the Insane and Feeble-minded (1906); Benevolent Institutions (1905); Paupers in Almshouses (1906); Prisoners and Juvenile Delinquents (1907); and a Report on Mines and Quarries (1902). A Census of Religious Bodies, continuing the volume on the Statistics of Churches, published as a part of the Census of 1890, is now in progress. All these inquiries have been heretofore taken as part of a decennial census.

It is unnecessary to mention here also the large number of additional inquiries which have been ordered by Congress and assigned to the Census Office. Investigations of this nature are frequently needed; and, had there been no permanent Census Office, they probably would have been undertaken by the Bureau of Labor or some other agency.

One of the

The bill recently reported by the Committee on the Census of the House of Representatives follows in all essentials the original changes introduced by the law of 1899. Like that law, it limits the enumeration strictly to four subjects, relegating other decennial inquiries to the interval between the censuses. inquiries of 1899, that into deaths in the general population, has been dropped. The reasons are twofold. Fifty years of experiment from 1850 to 1900 has established with certainty the fact that not more than seven-tenths of all the deaths which occur in a given community during a year can be obtained by enumerators asking at the close of that year of each family a report of any deaths which had occurred therein during the preceding twelve months. Tables so imperfect have very little statistical or medical value, and are constantly misunderstood by the public. So wide a margin of error in one branch of the census tends also to discredit more accurate results reached in other divisions. Furthermore, the Federal Census, acting in co-operation with the states and cities having trustworthy local registers of deaths, is now annually publishing returns for about one-half the population of the country, which probably cover 95 per cent. of all deaths occurring within that region,