The history of statistics : their development and progress in many countries

Front matter

.0^

,NC'

^0-

BOOK I Historical and Commemorative Addresses

THE AMERICAN STATISTICAL ASSOCIATION, 1839-1914 By John Koren, President

SEVENTY FIVE YEARS OF PROGRESS IN STATISTICS: The Outlook FOR THE Future 15

By S. N. D. North, LL.D., Assistant Secretary and Statistician of the Carnegie Endowment for International Peace; Ex-President of the American Statistical Association; Former Director of the United States Census

BOOK II History and Development of Official Statistics IN Many Countries

By George Handley Knibbs, C.M.G., F.S.S., Honorary Member American Statistical Association, Member U.S., etc.

By Dr. Robert Meyer, Privy Councillor, Former Minister of Finance, late President of the Imperial and Royal Central Statistical Commis-

By Dr. Armand Julin, Director-General of the Belgian Labor Bureau,

Member of the International Statistical Institute

By Ernest H. Godfrey, F.S.S., Member of the International Statistical

Institute; Editor, Census and Statistics Office, Ottawa, Canada

VI CONTENTS

By Adolf Jensen, Chief of the Statistical Department of Denmark

By Fernand Faure, Professor of Law at the University of Paris, Member of the Central Statistical Commission

GERMANY 331

By Dr. Eugene Wurzburger, Privy Councillor, Director of the Royal Statistical State Office in Dresden

By Sir Athelstane Baines, C.S.L, Ex-President of the Royal Statistical Society

By Dr. Ladislaus von Budat, Ministerial-Sektionsrat

INDIA 415

By Sir Athelstane Baines, C.S.L, Ex-President of the Royal Statistical Society

ITALY*

By C. A. AscHiERi, Inspector-General of Statistics

JAPAN*

By S. Takarabe, Professor of Statistics at the University of Tokyo

By Dr C. a. Verrijn Stuart, Professor of Statistics and Economics at the University of Groningen, Chairman of the Central Commission for Statistics

By A. N. KiAER, Chief and Director of the Statistical Bureau, 1867-1913

By Dr. A. Kaufmann, Professor of Statistics at Petrograd

* For the countries marked with an * the articles were promised by the authors named, but their work was interrupted by the war and the articles have not arrived in time to be included in the volume.

CONTENTS vu

By Dr. Edvard Arosenius, First Actuary of the Central Statistical Bureau

FEDERAL STATISTICS. By John Cummings, Ph.D., Statistician,

United States Bureau of the Census

STATE STATISTICS. By Charles F. Gettemt, Director, Mass- ^^ ^ ackusetts Bureau of Statistics

W

Introduction

The motive that inspired this volume was to mark the Seventy fifth Anniversary of the American Statistical Association by a serviceable contribution to our knowledge of statistics. The opportunity was found in preparing a history of the development and organization of official statistics throughout the world. By the aid of certain commemorative publications and other scattered documents one could piece together more or less complete sketches of the status of statistics in this or that country covering various periods of time; for most countries comprehensive and authoritative statements were lacking. The story had not yet been written as a whole and in detail how different nations have systematized information about themselves, what statistical means they employed, and what has been accomplished to the end that national forces and resources might be placed under wise direction and legislation.

The history of the development and organization of official statistics is not a barren record of steps in a scientific process of dealing with facts, but of efforts to get a working knowledge about the fundamental elements in the life of a country — the population, its environment, and its manifold economic and social relations. By taking measure of these elements, statistics reveal the condition of growth and trend in every direction and set out the milestones for the guidance of the administrator and legislator. But the manner in which the science of statistics has been applied also carries many practical lessons, for out of the abundant experience of others the rest of us must learn. Although the official statisticians of the different countries in the main are occupied with the same problems, the conditions under which they labor vary according to the tradition, experience, recognized needs and legislation of each country, out of which have grown the present forms of statistical organiza-

X MEMORIAL VOLUME

tion. No two countries have an identical programme of statistical work, nor the same agencies for carrying it out. The common aim has been diversely pursued, sometimes disappointingly and again with marvelous success. To tell the story of it all is the purpose of our Memorial Volume. Happily it is told by those who know it best, who in large part have lived it, and whose authority illuminates every page. The foremost statisticians of many lands were induced to write these memoirs to meet a need of so great an importance that not even the Great War could obliterate it.

The preliminary arrangements for the volume were made at the end of 1913 and the beginning of the following year. It was hoped to publish by the close of 1914. Some of the manuscripts were at hand before the outbreak of the war, and among them, the history of the statistics of Belgium, written by Dr. Armand Julin, reached the editor shortly prior to the fateful August 4, 1914. But several of the most important memoirs were still uncompleted, and the world was at war. After much anxious correspondence a Russian collaborator was secured, in Professor A. Kauf mann of Petrograd, but he could not finish his exhaustive study until the second year of the war. Dr. Eugene Wurzburger, Director of the State Statistical Department of Saxony, had only written a few pages of the history of German statistics when the exigencies of the war compelled him to lay it aside and not until 1916 could he conclude the difficult task. M. Fernand Faure, Professor of Law at the University of Paris and editor of the Revue Politique et Parlementaire, who had undertaken to prepare the memoir for France, was obliged to postpone it from time to time at the call of other duties. For most of the important belligerent countries the expected contributions were ready a year ago ; but it was resolved not to publish until France could be worthily represented. An appeal was finally made to His Excellency M. Jusserand, the ambassador of France to the United States, who graciously promised to intercede with his government so that Professor

INTRODUCTION Xl

Faure might be given leisure to assemble his rich material. This has been done in an extensive memoir of rare merit. Also for some of the lesser countries were the studies completed amid the distractions of the Great War.

Probably no other international work has been prepared under greater diflaculties, and none of more permanent value. It has come into being during a world conflict; it is without a counterpart in any language; it could not now be duplicated; and it marks a distinct era in official statistics which the war brought to a close. From this world-wide disturbance a new order of things will emerge that is certain profoundly to affect the future development of statistical work. Thus the volume in a special sense becomes an invaluable historical document telling the ways and means whereby civilized countries so far have taken measure of their material and social conditions — a monument to a bygone age.

The practical usefulness of these memoirs the reader must search out for himself. Only this may be said: There is nowhere evidence of contentment with past performances in the field of statistics; the shortcomings of most forms of statistical organization are cheerfully recognized; there is manifest desire, even expectancy, of richer service under better auspices; and the need is emphasized of cooperative effort so that the statistician's dream may be realized of international comparability in many statistical undertakings. For long years the masters of our science have striven toward this goal. Must they henceforth go their several ways because those who would stand together for the time have been parted? Perhaps this Memorial Volume may in slight measure serve to strengthen the international bonds and reknit those that have been severed, since it indicates so clearly that we cannot attain the highest ends of statistics except through mutual understanding and helpfulness.

To the many collaborators, some of whom the editor has unblushingly importuned at all seasons to complete their promised tasks, the American Statistical Association begs to

Xll extend its lasting gratitude. The volume is a memorial to the dead and living among them as well as to our Association. Thanks are also due to the members who have assisted in preparing translations of articles and given other support.

J. K.

January 1st, 1918.

BOOK I

HISTORICAL AND COMMEMORATIVE ADDRESSES

THE AMERICAN STATISTICAL ASSOCIATION

1839-1914

By John Koren

President, 1913 and 191 Ji- Seventy fifth Anniversary, February 13, 1914

A page of the unpretentious volume containing the first proceedings of our society bears this legend: "Here close the records of the American Statistical Association for the first quarter of a century." A special committee which had been appointed to consider the advisability of a formal celebration of the event, reported it to be "inexpedient." Did modesty deter them, or perhaps the overhanging shadows of the Civil War.^^ The records do not say. We of a later day know that these pathfinders in a statistical wilderness were not at all disheartened as they looked back upon their labors of twenty ^ve years. Indeed, there seems to have been something prophetic in the words, "Here close the records of the American Statistical Association for the first quarter of a century." Who wrote them looked forward to other milestones in the history of our organization, to new and perhaps ampler records of achievement; and we are gathered to bear testimony that his faith did not belie him.

That small company of men, meeting in 1839 at No. 15 Cornhill, Boston, to organize a statistical society, built securely upon a broad foundation. "The objects of the society," says the first constitution, "shall be to collect, preserve and diffuse statistical information in the different departments of human knowledge." How these objects were to be reached is elaborately set forth in the by-laws, no less than thirty three in number. They provide, among other things, that "the operations of this Association shall principally be directed to the statistics of the United States; and they shall be as general and as extensive as possible and not confined to any particular part of the country. . . . The labors of the Association may embrace all subjects." The by-laws also planned for a statistical library and stated that "efforts to diffuse statistical information shall be made by printing and publishing circulars, reports, a periodical or occasional volume." The very first committee appointed after organization had been perfected was one to have charge of eventual publications.

From the outset membership in the Association was considered a serious affair. On being elected, the fellows subscribed to a solemn "obligation" to conform to the constitution and by-laws. It was made "the duty of every fellow to prepare at least one article a year on some statistical subject which shall be at the disposal of the publishing committee." And to add greater dignity to fellowship, it was stipulated that "a diploma, in a form prescribed by the board of directors, signed by the president and recording secretary, accompanied with the seal of the Association, shall be given to every member."

Thus seventy five years ago the fair enterprise was launched by a little band of enthusiasts who labored diligently to inform themselves and others, and who, looking forward with rare courage to the things to be, rather than to the present, wrote aboult the Science of Statistics, always spelling it with a capital S. At that time the professional statistician was not, and the statistical output almost negligible.

Even in its early stages the Association lacked that parochial flavor which some have imputed to it. It was not preoccupied by local affairs or wasting time in mutual selfadmiration over Massachusetts men and things. At the very outset contact with a larger world was sought through its foreign and corresponding secretaries, and within a year it counted corresponding members in about a dozen states. At the first quarterly meeting it reached out beyond the United States by electing foreign members, the fir^t of whom was the foremost statistician of his day, Adolphe Que tele t.

There is something touching in the fidelity with which the founders of our Association sought to work out its objects. Year by year the same small group, usually under a dozen, regularly attended the four quarterly meetings, encouraging each other to original effort in statistical work and garnering diligently helpful knowledge from various sources. Nothing testifies more eloquently to the statistical poverty of the time than the publications collected as the beginnings of a library. Among the works donated we find mention of Pitkin's "Statistical View," Hazard's "Statistical Register," "Bills of Mortality of Philadelphia from 1820 to 1839," by Dr. Henry Bond, the few volumes of the United States Census, schedules for taking the census, statistical forms and questions used by the Statistical Society of London, forms used in France for the registration of births, marriages and deaths, unnamed statistical tables of Massachusetts by A. W. B. Peabody, etc. Then there was the imperfect material found in local, state and federal reports of all kinds which were collected in large numbers. Many non-statistical works were presented to the library, sometimes by the authors. Thus Emerson gave his lectures and Prescott various volumes.

A feature of these early meetings was the reading of communications of different sorts. Mention is made of letters received from President Martin Van Buren, Thomas L. Winthrop, John Pickering, Benjamin F. Butler, LL.D. (of New York), Levi Woodbury and William Prescott, all of them honorary members; and from Caleb Cushing, Lewis Cass, Governor Plummer, Thomas Lawson, Surgeon-General of the Army. Following the reading of letters came the statistical essay. The first one recorded was by Lemuel Shattuck on statistics of Saxony. The early addresses covered such subjects as statistics of pauperism, of crime, of immigration, of the Massachusetts population, etc. Some bore curious titles. Thus the indefatigable long-time recording secretary of the Association, Joseph B. Felt, presented statistics of the number and kind of carriages in Massachusetts in 1756. Another gave statistics of the funeral charges in the interment of Governor Winslow in 1680, and once, when there must have been a dearth of statistical material, someone read a paper on a tornado that had disturbed the calm of Medford.

Not all who dived into the uncharted statistical seas brought up a perfect pearl. Some of us fail to do so now, and those intrepid founders of our Association were indeed rari nantes in gurgite vasto. Yet solid contributions to knowledge were made. Within ten years there had been read before the Association more than thirty addresses upon a wide range of topics. The question of publication soon became pressing. Although the first volume of the Association was not published until 1847, the first part of it appeared in 1843, the second in 1845 and the last in 1847. "They give," says Dr. Samuel A. Green, *'a large amount of original statistical matter, and on every page show signs of patient research and thorough work." *

But the Association found other work to hand than that of reading essays, and soon set about the greater task of improving the sources of statistical knowledge. Naturally enough, attention was first turned to vital statistics. Already in 1840 a committee was appointed to "lay the subject of registration before the General Court," and in the following year a committee was formed to report on a plan of registration. Evidently, the legislature was not very tractable, for it was memorialized on the subject of vital statistics at intervals for many years. Other things for which the Association sought legislative authority during the first twenty five years of its existence were: The appointment of a sanitary commission (1847-8); for improved hospital statistics; for returns to the Secretary of State of pauperism, crime and immigration (1850); for a decennial census of the commonwealth beginning 1855 (1850); for the establishment of a state board of health, to have charge of registration

* An account of the collections of the American Statistical Association by the Hon. Saml. A. Green, M.D., Librarian Massachusetts Historical Society, new series of publications No. 31.

and of the census, and several other matters. For long years afterwards the Association continued its activity for the betterment of Massachusetts statistics; and to its efforts is due in large part that this commonwealth has taken foremost rank in statistical work.

Meanwhile, the Association had sought larger fields. From the beginning it became closely identified with the affairs of the United States census. In 1844 it petitioned Congress that the Sixth Census "be revised and a new and accurate copy be published." As the time for taking the Seventh Census approached (1848), a committee was appointed to prepare and submit plans for it and to memorialize Congress "to take measures to render the Census more accurate." It was at the suggestion of Lemuel Shattuck, a charter member of the Association, and N. Capen, another member, that Congress created a census board charged with the planning for the Census of 1850. Edward C. Lunt, in his History of the United States Census, records that "several eminent statisticians," — Mr. Shattuck, Dr. Chickering, Mr. Capen and Dr. Jarvis (then President of the Association), '*were invited to Washington for consultation with the new board." Indeed, we know that Lemuel Shattuck, who had gained large practical experience in compiling a census of Boston for the year 1845 and in other ways, prepared most of the schedules adopted. Dr. Jarvis became closely identified with the Eighth Census, that of 1860, for he wrote the section on vital statistics. He complained (it sounds so reasonable) that the material furnished him was often "imperfect, inconsistent or unreasonable." The same indefatigable Dr. Jarvis also rendered valuable services to the Census of 1870. These men set high standards, insisting upon it that those entrusted with the oversight of important statistical work "should be appointed, not for their political opinions but for their scientific attainments and knowledge," and they were unsparing in their criticism of a faulty output.

But to realize how closely this Association has been affiliated throughout its entire history with the United States

Census, it suffices to record the names of some of the men at the head of the Censuses: John B. D. DeBow, superintendent of the Seventh, was a corresponding member of our Association; General Walker, its president from 1883-1897, had charge of the Ninth and initiated the Tenth; Colonel Wright, our president from 1897-1909, finished the Eleventh; S. N. D. North, another president, was the first director of the permanent census office; and E. Dana Durand, for several years one of our vice-presidents, was director of the Thirteenth Census. Space forbids the mention of all officers and members who in the remote or recent past have rendered service to the great office of the census.

Other branches of the federal government occasionally sought the help of the Association. Particularly was this true of the Treasury Department. In 1845 its secretary requested the Association to aid him in preparing his report to the next Congress. In later years the Treasury Department appears to have been in frequent communication with the Association and several of the secretaries have been among our members. The Commission on Education also drew upon the services of the Association.

Although the Association from the beginning had sought touch with European statistical affairs, it was not until 1860 that an opportunity presented itself for participation in an International Statistical Congress, the one held in London that year, and to which the Association fitly sent Dr. Jarvis as its representative. Some time later a letter he had written from London at that time was read at a meeting and the records state that in it he "detailed the action of Judge Longstreet of South Carolina, the delegate from the United States, and Mr. Dallas, the American Minister, relative to a remark of Lord Brougham construed to be an insult to our country, — Dallas and Longstreet seemed at the time to consider slavery of supreme importance."

The next International Statistical Congress, at Berlin, 1865, was attended by E. B. Eliott as the delegate from the Association. He, together with Dr. Jarvis, was delegated to the International Statistical Congress held at St. Petersburg. Dr. Jar vis apparently did not go; the Hon. E. H. Derby, another member, seems to have taken his place. It is not recorded that the Association was officially represented at any other statistical meeting abroad until 1885, when General Walker attended the conferences at Paris and London, which resulted in the formation of the International Statistical Institute. It is interesting to note that as delegates returned from these international meetings, they were especially requested to report upon the progress of statistical science abroad. Another means of communication with European statistical interests was through the honorary and foreign members like Rawson W. Rawson, John Bowring, William Farr, some of whom from time to time sent in communications, and through statistical organizations, particularly the Royal Statistical Society of England, whose publications are frequently referred to.

While thus cultivating the larger fields, the Association adhered to its routine of quarterly meetings. Even during the dark years of the Civil War they did not suffer a single break, but much attention was then paid to military statistics, army hospitals, sanitation and similar subjects. The president, Dr. Jarvis, and several other members did notable service in sanitation work. At this period a new and simple constitution was adopted and the by-laws were reduced from thirty three to eight articles. The range of topics upon which addresses were read broadened in several directions. Statistics of commerce, finance, agriculture, taxation, shared attention with those of population, immigration, mortality, etc. Occasionally consideration was given to statistical method. One matter upon which the Association declared itself deserves special mention. In 1868 the Association received an elaborate report on the metrical system of weights and measures, prepared by two of its members, and not only went on record as favoring its adoption by the people of the United States but voted to petition Congress on the subject and to ask the legislature to pass a statute requiring the metrical system to be taught in the pubHc schools. The action seems to have attracted much attention at the time.

The many-sided activities of the Association in the early periods of its life is the more surprising when one considers the limited membership. Until a time easily within our memory, little effort was made to get new members. Perhaps one reason for this was that a large constituency was not needed since little was published. The founding of a statistical journal was frequently discussed, but not even preliminary steps were taken. In 1869 the second volume of the collections was ordered printed — twenty two years after the first. Unfortunately, many delays occurred; and in 1873 President Jar vis reported that a part of the printed pages of this second volume, which had not been bound, were destroyed in the great Boston fire.

A question that on many occasions agitated the Association and probably gave rise to long discussions was that of the library. It had grown apace through "donations," Dr. Jarvis alone having during his lifetime given it more than 600 volumes. Dr. George C. Shattuck had been another constant benefactor also through contributions of money. How to house a collection which in 1875 numbered about 2,500 volumes had long been a perplexity. Later on it received large accessions, and as the Association had no rooms of its own, the difficulty grew. Finally it was transferred to the building of the Institute of Technology and in 1899 found its permanent resting place in the Boston Public Library. Before that time it was officially designated as the "Jarvis Statistical Library."

Before sketching the last twenty five years of the history of the Association, let me dwell for a moment on the personality of some of its early leaders. The brunt of work during the first years fell to Joseph B. Felt, long the recording secretary. The first president, the Hon. Richard Fletcher, seems to have been little more than an amiable figurehead. Mr. Felt appears to have possessed much curious learning with a decided bent for historical research. He was an indefatigable worker and contributed more than his share of essays. Of Lemuel Shattuck, another of the founders, we already know that he has been characterized as an eminent statistician. He deserved the name. His census of Boston is a remarkable piece of work, and the sanitary survey of Massachusetts prepared under his direction is a classic. He was not a frequent essayist, but gave unstinted service in a larger way. The other Shattuck, Dr. George C, president from 1846 to 1852, merits a warm tribute for his long and unflagging devotion to the Association. During life he gave liberally to it and he remembered it in his will.

He was followed by Dr. Edward Jarvis, for thirty one years president, who stands out as the most remarkable figure among his contemporaries in the Association. I can do no better than to quote from the resolution adopted at his retirement as president. The Association then recorded the "Earnest expression of the manifold service he has rendered the Association by the able manner in which he has performed the duties of his office, by the ability with which he has represented the society at home and abroad, by the many valuable papers he has prepxred for us and read at our meetings, and by the assiduous and unwearied zeal with which he has labored at all times and in all ways for the advancement of the interests of the Association and of statistical knowledge."

It may truly be said that no human interest was foreign to Dr. Jarvis. In his speeches and writings on statistical subjects he traversed the widest fields and illuminated many ^' an unexplored corner. His chief interest lay, however, in vital statistics, for the improvement of which he labored unceasingly. For thirty one years he carried the heaviest load for the Association. It is on record that during this time he prepared more than thirty five papers and addresses for the meetings. Upon his retirement in 1882 the Association paid him the greatest tribute that lay within its gift by making him president emeritus.

Another remarkable figure was that of S. B. Eliott, who s described in our records as an "insurance actuary and calculator," of "accurate discipline, mathematical learning and laborious industry." He later became attached to the United States Treasury Department. At the request of the Prussian government he constructed the first life table for that country (published in the proceedings of the American Association for the Advancement of Science, Albany, 1856). He later constructed a second life table for the same government. He made a number of studies of comparative mortality for different countries which led him to the conclusion (in 1858) that "the mortality of the United States is such as to make us dependent on immigration for permanent increase."

Among men of the early days who deserve to be singled out for special tribute for their services are these: Hon. Amasa Walker (the father of Francis A.), for fifteen years a vice-president and frequent essayist; Joseph E. Worcester, foreign and corresponding secretary for twenty six years; John Ward Dean, in recognition of whose long service as recording secretary he was made a life member; J. Wingate Thornton, another recording secretary and contributor of statistical papers.

In the third decade of Dr. Jarvis' administration, two strong men became members who have left a lasting mark upon the Association. First came Carroll D. Wright, who had taken charge of the Bureau of Labor Statistics, and a little later Francis A. Walker, then professor at Yale University. Both brought new interests to the Association and gave added impetus to the efforts for improving the statistical service, both local and national. Among the honorary members elected at this time, the name of Florence Nightingale stands out. In acknowledgment she remembered the Association with some sanitary reports for India and a pamphlet of her own authorship (Life or Death in India) .

With the presidency of Francis A. Walker began a period distinguished, among other things, for the accession to membership of many university teachers who as a class had hitherto shown little interest in the Association. No doubt Walker himself proved a powerful magnet to the academicians, and the need of teaching statistics had begun to be recognized. But the great event in President Walker's administration was the founding of the so-called new series of our publications in 1888. It is strange that our records should remain silent about so momentous an undertaking — our chief permanent contribution to statistics and the lasting mem_orial of the activities of our Association. The new venture demanded the support of a larger membership and it was sought; the response was immediate. Soon the Association had expanded from a local to a national group. In a few years (1895) it counted more than five hundred members, high-water m^ark for that time being reached a little later. The heavy work in those days fell to Dr. Davis R. Dewey, for twelve years recording secretary and editor of the publications, to whom the Association owes an unforgettable debt of gratitude for efficient and ungrudging service. As if to emphasize the larger life upon which the Association had entered, it was decided to have quarterly meetings at Washington, and one was held there in 1896.

Of the general activities of the Association under President Walker's eminent and devoted leadership, the printed volumes afford testimony I need not supplement. Upon his death in 1897, Carroll D. Wright, who had served as vice-president, recording secretary and librarian, lent the prestige of his rare personality and achievements as statistician to the office of president. Under him the Association continued to flourish for some years. Later some lean times set in, a period of temporary retrogression that seems unavoidable in the history of every organization. Membership declined and the meetings dwindled to an annual meagerly attended affair. Some of us here remember well those days of disheartenment. The reaction did not last long; and, happily. Colonel Wright lived to see the Association once more in the ascendancy and entering upon new and better days. In his last address to us (1908), he painted large the spreading field of usefulness waiting to be occupied by the Association, confident in its strength and loyalty to ^ high ideals.