CORNELL UNIVERSITY LIBRARY Date Due IV uw™f' <^- — ^ W^T MA1U7' m± jal DPCT*-* £W* ^Mll ~&1S/9f HQ 814.W49 e " Un,Ver " ,y Ubrary The divorce problem a study in statistic 3 1924 021 866 706 The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924021866706 STUDIES IN HISTORY, ECONOMICS AND PUBLIC LAW. EDITED BY THE UNIVERSITY FACULTY OF POLITICAL SCIENCE OF COLUMBIA COLLEGE. Volume I.] [Number 1. THE DIVORCE PROBLEM. A 8TUDY IN STATISTICS. BY Walter Francis Willcox, Ph.D., SEUGMAN FELLOW IN POLITICAL SCIENCE, COLUMBIA COLLEGE. INSTRUCTOR IN PHILOSOPHY, CORNELL UNIVERSITY. New York. 1891. lt/\~ PREFACE. The final form of this monograph is the result of a conver- sion. My study of divorce was commenced when fresh from the reading of philosophy in Germany, and a month or more passed in turning the leaves of Trendelenburg, Bluntschli, Stahl and the whole line of "Naturrecht" theorizers. Nothing was found to shake the conviction with which I started, that the policy of the Catholic church, refusing remarriage in all cases, is the ideal one for a state to adopt. Then I stumbled upon Bertillon's Etude Dimographique du Divorce and, unde- terred by the columns of figures, read and reread it. My eyes were opened and, deserting the high a priori road of laying down what marriage and divorce ought to be, I betook myself to a patient examination of Mr. Wright's Report in the effort to understand what they are. My conclusions are contained in the following pages. In their present form, therefore, they are based on two books; their method is derived from Bertillon their data from Wright, and a critic must have keen eyes to detect in them any influence of the first six weeks' reading. If a similar revolution should be started in the mind of any reader by the facts here recorded, I shall be most amply re- paid. It is a cause of regret that I have been compelled so often to differ from, or criticise the results of, the able statistician at the head of the Labor Department. No one can value more highly the work Mr. Wright has carried to success in the face of numerous difficulties. The proof of my admiration, however, must be found in the weeks of toil I have profitably spent over the book rather than in any words of empty praise. (Hi) iv PREFACE. Some errors would be almost inevitable in reviewing so com- plicated a subject as the divorce legislation of the various states and the changes it has undergone in the past twenty years. If any such have been made by Mr. Wright, and as these pages are going through the press I have received some reason to be- lieve that in the case of Vermont (§27) they have been, they will indirectly affect my results. In such cases my criticism would apply to. Mr. Wright's evidence for the influence of legislation, and not to the facts. There is no reason to doubt, however, the correctness of practically all the statements of fact in the Report. For this reason, and because their verification would have been in many cases impracticable, I have not at- tempted it in any case. In conclusion, I desire to express thanks for their courteous assistance to Dr. S. W. Dike, Secretary of the National Di- vorce Reform League, to the Librarians of Columbia College and the Massachusetts State Library and, above all, for fre- quent and full replies to all letters of inquiry, to the Hon. Carroll D. Wright, U. S. Commissioner bf Labor. W. F. W. Malden, Mass., March, 1891. TABLE OF CONTENTS. PAQE S i. Introduction 9 Part I. General Statistics of Divorce. ' § 2. Comparison between the United States and other countries 11 y§ 3. Rate of increase . . 12 §§4,5. Two possibilities of error 13-14 § 6. Increase with reference to courts granting divorce . . 15 § 7. Increase with reference to the population 16 § 8. Increase with reference to married couples 16 § 9. Increase with reference to its possible outcome ... 19 § 10. Increase among the negroes 21 § n. Increase in the "several states 22 § 12. Duration of marriage before divorce 23 § 13. Remarriage after divorce 26 Distribution of Divorce : § 14. Between natives and foreign born 28 § 15. Between Catholics and non-Catholics 29 § 16. Between negroes and whites 29 § 17. Between city and country 32 § 18. Between couples with children and without 33 § 19. As granted to husband or wife 34 § 20. Among the several states 37 Part II. Influence of Legislation on Divorce. v-§ 21. Problem and method 41 t— § 22. Legal causes of divorce 41 w§ 23. Comparison of number of causes with divorce fate . . 44 (v) vi CONTENTS. PAGE § 24. Changes of law 47 — ■§ 25. Introducing causes 48 ■HI 26. Repealing causes 49 -§27. Modifying causes or procedure 50 § 28. Of greater complexity 52 ^29. More radical European changes 55 § 30. Affecting the expense of divorce 58 §31. Introducing uniformity in Switzerland 59 § 32. Restricting marriage 60 § 33. Summary of results 61 Part III. Causes and Remedy. § 34. Basis and character of conclusions 62 § 35. Two conceptions of marriage law 62 § 36. The popularization of law 63 § 37. Laxity in changing and administering the law .... 64 § 38. Age of marriage 65 § 39. The emancipation of women 66 § 40. Growth of cities' 68 § 41. Increase of industrialism 68 § 42. The spread of discontent 70 § 43. Two ideals of the family 71 § 44. The remedy 72 LIST OF TABLES. PAGE I. § 2. Statistics for the Christian world in 1885 .... 12 II. § 10. Comparative increase in white and black counties . 21 III. § 11. Decennial increase in each state 22 IV. § 1 2. Average length of marriage before divorce in each state 24 V. § 12. Increase according to duration of marriage ... 25 VI. § 13. Remarriage of dfvorced persons in Switzerland . 27 VII. § 16. Divorce in white and black counties 31 VIII. § 19. Percentage of divorces granted to wife 36 IX. § 20. Divorce rate of each state in 1870 and 1880 . . 37-38 X. § 22. Divorces classified by causes 43 XI. § 31. Rate of Swiss cantons, 1876-85, under uniform law 59 (vii) THE DIVORCE PROBLEM. A STUDY IN STATISTICS. § I. Introduction. The "Report on Marriage and Divorce" transmitted to Con- gress two years ago by the Commissioner of Labor, Hon. Carroll D. Wright, might better be entitled a " Report on Divorce." Statistics of marriage are conspicuously absent. «' Thoroughly incomplete and unsatisfactory" is the Commis- sioner's own characterization of that part of his work. Almost its only value is to reflect and reveal the wretched condition of marriage records in most of our states and territories. But the bulky volume is a mine of information on the subject of divorce in this and foreign countries. Like other mines, how- ever, it does not carry its ore on the surface; it needs to be worked. The valuable introductory chapters of Mr. Wright have by no means given an exhaustive interpretation of its figures, while some of his conclusions, indeed, are questionable or erroneous. This will be shown at length in the course of the following argument. The discussion of divorce maintained in the periodical press since the publication of this Report, has done little more than arouse interest and diffuse a vague conviction that something must be done, perhaps in this way occasioning a move recently made by the New York Legislature, under the inspiration of Governor Hill. A Commission of three has been appointed (9) 10 THE DIVORCE PROBLEM. in that state to promote uniformity of legislation in our vari- ous jurisdictions, and has taken the conflicting laws of marriage and divorce under consideration in the effort to harmonize their differences. A commendable object, surely; but sup- pose it accomplished, how much good would be done? Would the divorce movement be checked or even appreciably affected ? The publication of the Report and the appointment of the Commission are the most important of recent steps towards divorce reform; but the relation of the two, the bearing of the former on the latter, has not been recognized. In reality, Mr. Wright's volume contains an answer to the question : What can such a commission accomplish for the reform ? But this answer does not lie on the surface, nor has it been read truly and clearly in the analysis of the figures. To reach it we must mine into the tables, re-arrange and re-interpret them. The direct object of this paper is to determine the influence of legislation on divorce by conclusions drawn from the raw ma- terial of figures now offered for study. The question must be attacked by slow and indirect approaches, starting with a more exact ascertainment of the nature and dimensions of the prob- lem. Accordingly, the first part will be occupied with a deter- mination of its general statistical phases ; the second, with a study of the effects of legislation in this and foreign countries ; and the last, with some conclusions upon the causes and true remedy for divorce. PART I. GENERAL STATISTICS OF DIVORCE. § 2. Comparison between the United States and Other Coun- tries. The statistics of foreign countries presented in the excellent Appendix to the Report make such a comparison possible. The largest absolute numbers are reported from Germany and France ; each country has about one-fourth as many divorces as the United States. The -largest numbers relative to popu- lation are found in Switzerland and Denmark, whose divorce rate is but little better than the average for our whole country; six Swiss cantons, indeed, show as high a rate as the New Eng- land states. The smallest numbers relative to population are found in Great Britain and Ireland, the British colonies, Can- ada and Australia, Norway, Russia and Italy. On summing up the divorces and separations in all the countries whose records are given, it appears that more are probably granted in the United States than in all the rest of the Christian world, Protestant, Catholic and Greek; that is, more than in all Europe, outside the Balkan peninsula, all civilized Australia and Amer- ica. On account of the insufficiency of data this result can be stated only as a probability. Returns are given, however, for all Christian Europe except Spain, Portugal and Greece. Outside of Europe, the Report gives statistics for Canada only, and to these I may add some figures for Victoria* and New South Wales f. In a few instances the data for 1885 are not given, and must be estimated from those of earlier years. * Victorian Year Book 1880-81. t Handbook of New South Wales Statistics for 1887. (11) j 2 THE DIVORCE PROBLEM. The following table summarizes the divorces granted in 1885 in the various countries of the Christian world, from which either statistics or an estimate based on statistics can be obtained. Canada 12 Great Britain 1 ~ and Ireland J " ' France 6,245 TABLE I. Holland 339 Denmark* 635 Norway * 68 Sweden 229 Russia '1789 Australiaf 100 Italy 556 Germany 6,161 Switzerland 920 Austria 1,718 Belgium 290 Roumania * . . . . 541 Total 20,111 United States 23,472 Excess 3t36' It is very doubtful whether the number of divorces and sepa- rations in Spain, Portugal, Greece, Mexico, Central America and South America, all told, would equal this excess of over 3,300. But in the absence of direct evidence the reader must form his own opinion on that point from an examination of the table. So far as statistics are published we may safely go ; and we may say that the number of divorces in the United States is considerably in excess of the number reported from all the rest of the world. § 3. Rate of Increase. The divorce question, however, is fundamentally a problem, not in social statics, but in social dynamics. The number of divorces granted in a single year, either in one country or in the world, is no measure of its importance. It is a rising tide and derives its main significance not from its level at any one moment, but from the rapidity of its flow. ' The tide is rising steadily all the world over, but nowhere is it so highj nowhere does it rise so fast, as in these United States. Alarm- ists exclaim that it betokens the ruin of our whole social fabric> and free-lovers exult that the world is advancing towards the light. To the optimist it is merely a temporary set-back, an eddy in the sweep of our triumphant democracy ; but to the * Estimated on the basis of returns for previous years, f Estimated on the basis of partial returns. THE DIVORCE PROBLEM. I3 sociologist it is a phenomenon to be neither approved nor condemned, only analyzed and explained. In attempting to estimate from the statistics the rate of in- crease of divorce we are met on the threshold by two possi- bilities of error. § 4. The number of divorces is not reported with absolute accuracy, for the records have not been preserved intact in all parts of the country. The Chicago fire destroyed the records of Cook county, the Cincinnati court-house fire those of Ham- ilton county; and these two are but conspicuous and familiar instances of what has happened in ninety-eight counties* in the course of the twenty years. These counties are charged with 39 divorces for the first year of the twenty, and 1,768 for the last, an increase of 1,729. The other 2,526 courts with complete records show an increase of 13,869. That is, the former, with one -twenty-fifth the population of the country, are apparently responsible for one-ninth the increase. To eliminate the obvious error it is neccessary to know the true rate in these counties. A simple method of finding it would be to apply to them the average rate of increase in the rest of the country ; but a moment's thought will show this method to be inadmissible, because the increase of divorce is closely related to the increase of population, and the latter is much more rapid in these counties than elsewhere. From 1870 to 1880 they grew in population 47 per cent, and the rest of the country 29 per cent. Still further, divorces increase in large cities like Chicago and Cincinnati much more rapidly than population ; and the same is true of the southern and south- western states, which contain a large majority of the other 96 counties. Therefore it cannot be far wrong to estimate that the increase in these counties has been twice as rapid as the average for the rest of the country. On that assumption *The records of 160 counties are imperfect, but in only ninety-eight have the divorce records been so injured as to affect the accuracy of the total increase for the twenty years. 1 4 THE DIVORCE PROBLEM. their divorces in 1867 numbered 344, instead of 39, and the total increase in the country for 20 years was 15,293, instead of 15,598. The exaggeration of increase due to this cause is thus about two per cent. § 5. The former error referred to the success with which the records once made have been preserved. A second may lurk in a neglect to make the records complete. Some divorces may have been granted, but not entered at all. In the very nature of the case, no proof of such an error, and, a fortiori, no measure of its magnitude, can be derived from the Report. One slight indication is all I have discovered. It would be supposed that the court roll would indicate the cause for which the decree is granted with the same uniformity with which the records of criminal courts state the ground of sen- tence. Yet 10,315 decrees were reported to Washington as issued for "cause unknown." When the rolls show such neg- ligence as this, when even some large states like Pennsylvania, Indiana and Texas record from ten to fifteen per cent, of their divorces without mention of any cause, it may be assumed that occasionally a little greater neglect has omitted the entry alto- gether. Admitting this to be true, the proportion of such unrecorded divorces was probably greater in 1867 than in 1886, for the records have come to be more carefully kept. In 1867 4^ per cent, were reported as decreed, for cause un- known, but in 1886 the percentage had sunk to less than 2)4- Therefore, the error, assuming it to exist, would not only affect the absolute number for each year, but also accentuate the rate of increase, just as the imperfection of the census of 1 870 led to an overstatement of the growth in population in the southern states in the following decade. It is probable, then, that as the years went by an increasing proportion of the divorces granted were recorded ; and an in- creasing proportion of those recorded were preserved until Mr. Wright's investigation. These two errors will enter into the following calculation of the rate of increase and affect the THE DIVORCE PROBLEM. 1 5 results. The former might, perhaps, be eliminated, but only by a long and intricate computation for each case and the latter would still remain quite indeterminate. Therefore, a frank statement of them at the start must atone for their entire neglect in the subsequent discussion. § 6. Increase with reference to Courts Granting Divorce. Divorces are granted in the United States ordinarily by county courts. A few counties have two divorce courts. A somewhat large number of courts have jurisdiction over two counties. Therefore, the number of courts is-dosely depend- ent upon the number of counties, but somewhat smaller. In 1 886 there were about 2,734 counties and 2,624 courts. As the number of counties in 1867 was about 2,243, the number of courts may be computed approximately as 2,153. I n th at y ear they issued 9,937 divorces, or an average of .4.6 to each court. In 1886 the 2,624 courts granted 25,535 divorces, or an aver- age of 9.7 to each court. So the number of divorces grew more than twice as fast as the number of courts. But prob- ably not one of these courts is occupied exclusively with such suits, and their other business of all kinds must have in- creased rapidly in the twenty years. They have become more and more crowded with cases, and in danger of falling into 'arrears with their docket. Many of them have had to face the same difficult question that has recently vexed the national House of Representatives: How shall it reconcile its duty to transact the nation's increasing business with the seemingly incompatible duty of continuing a deliberative body? In this quandary too many judges have treated divorce cases as Con- gress has treated private pension bills. In both cases, there has been an apparent harmony of interests. In one the pen- sion applicant has wanted to obtain his,money, the Treasury to reduce the surplus and the Congress to save its time; in the other, parties to a divorce suit have been seldom at vari- ance in the single matter of desiring the decree, and the judge has often deemed it a matter of private concern, and saved his 1 6 THE DIVORCE PROBLEM. time for public affairs or suits involving a real conflict of in- terests. Nor has he been restrained so much as usual by the danger of appeal; that has been reduced to a minimum. § 7. Increase with reference to the Population. The census of 1870 is admitted to be inaccurate and that of 1890 is claimed to be so. Therefore it is difficult accurately to determine the rate at which the population has been increas- ing. Without turning aside to discuss the question, it may be assumed that the annual increase has been approximately 2^ per cent., and the decennial increase approximately 28 per cent. Of course, the increase of population has varied slightly from year to year, but nothing like so widely as the increase of divorces. The latter has swung between the extremes of a de- crease of .9 per cent, in one year, and an increase of 15.1 per cent, in another. But the average is 5.4 per cent, a year, or somewhat more than twice the increase of population. The second decade increased over the first 69.2 per cent. § 8. Increase with reference to Married Couples. The old law distinguished two kinds of death, natural and civil. Natural death was the end of life by disease, accident or otherwise ; civil death was its constructive or legal termin- ation, as by conviction of treason or admission to a monastery. Civil death is no longer recognized, but to this day a marriage may die either naturally, by the death of one or both of the partners, or civilly by a decree of divorce. Now, the death rate is found by comparing the annual num- ber of deaths with the number living; and, similarly, the divorce rate, or the civil death rate of marriages, is to be found by comparing the number of marriages dying annually by decree of court with the rtumber remaining alive. For determining the divorce rate, therefore, the number of married couples is as indispensable a datum as the number of divorces. Unfortu- nately in the United States the former is only a quaesiium; no national census has ever furnished this important information. THE DIVORCE PROBLEM. jy Therefore Mr. Wright has attempted to supply its place by an estimate. Several states in various parts of the country, Mas- sachusetts, Rhode Island, New York, Michigan and Kansas, have given in their state censuses the number of married couples. Mr. Wright has obtained the average of their results and extended it to the whole country, except a few frontier states. In this way the number of married couples has been determined to be about 18.9 per cent, of the population. With the fragmentary material obtainable, probably no better method of arriving at a result could be devised, and yet this is so inaccurate as to be almost useless. If the number of married couples be admitted to stand in a constant ratio to the. population, there is no advantage in comparing the divorces, with the former rather than the latter. But it is precisely be- cause the ratio is not constant that scientific statisticians insist that the comparison should be with the number of married couples, not with the total population. No fault could be found, with Mr. Wright's tables (pp. 148, 149, 160, 161,) if theic un- trustworthiness were stated. But, on the contrary, the incor- rect and misleading assertion* is made (p. 147) that, "in every instance in which it has been possible to test the estimated, number by the actual facts, it has been found to be either true: or within one-half of one per cent, of the truth." In reality the. ratio varies from 18.9 in New York (1875) to 19.9 in Michigan (1884), and from 17.1 in Michigan in 1854 to 19.9 in the same.- state in 1884. The variations shown by counties within the same state are much greater. In Massachusetts (1875), Berk- shire county had a ratio of 16.3 and Barnstable one of 22.5. In New York state (1875), New York county showed 17.2, and Cattaraugus 21.8. Obviously these variations in the ratio repre- sent variations more than five times as great in the number of * Here, and throughout this monograph, my criticisms are based on the first edition of the Report. As these pages are going through the press, I am informed by Mr. Wright that this assertion has been corrected for the forthcoming edition. My general criticism on his assumption that the ratio is constant remains, how- ever, unaffected by the change. 2 1 8 THE DIVOR CE PR OBLEM. maf'ned couples computed from them. Accordingly, among the instances in which I have found data for verification of the estimated number of married couples in a state, the maximum error has been, not .5 per cent, as stated, but 5.5 per cent, and when the same ratio is subsequently applied to New York city, the error rises to 9.9 per cent. Mr. Wright's tables, then, are not to be rejected, for nothing better can be substituted, but neither are they to be accepted without reserve. They are practically a comparison of divorces with population, thrown into the other form for convenience of statement and of com- parison with foreign statistics. In comparing divorces with married couples, the Report adopts the method of stating the number of couples to one divorce. This is like forming mortality tables from figures representing the number living to one death. The alternative method will be followed in this paper, and tables constructed by estimating the number of divorces to every 100,000 couples; the main advantage being that in such tables the numbers in- crease with the divorces, while by Mr. Wright's method the numbers decrease as divorces increase. It has already been observed (§ 7) that the annual number of divorces fluctuates widely ; and what is true of the country as a whole is more emphatically true of each individual state and territory. Hence, the figures for a single year offer too narrow a basis for accurate tables, and the only alternative is to take a mean of the numbers for consecutive years. This method will be adopted, and the number of divorces in 1870 uniformly computed by averaging the number for the seven years 1867- 1873, and similarly the number for 1880 will be found from the seven years 1 877-1 883. The treatment of Delaware in the Report is a glaring example of the error that may result from a neglect of this precaution. About 80 per cent, of the divorces in that state are granted by the legislature,* and the legislature meets biennially in the odd years. This explains * This information has been kindly furnished me by Mr. Wright. THE DIVORCE PROBLEM. I0 why the annual numbers vary from I in 1870 to 21 in 1869 and 21 in 1871, and from 5 in 1880 to 36 in 1879 and 20 in 1881. By comparing the numbers in 1870 and 1880 with the estimated number of married couples, Mr. Wright concludes (p. 148) that Delaware is more free from divorce than any state except South Carolina, which has no divorce law. The broader and fairer basis of comparison here adopted, relegates Delaware to the seventh place in 1870, and the fourth in 1880. (§ 20). As our attention is confined for the present to the country as a whole, the increase of divorce relative to married couples maybe stated in a single sentence. In 1870 there were 155. divorces to 100,000 couples; in 1880, 203. This indicates the degree in which the divorces are gaining on the population ; but in so abstract a form the full significance of the figures is not evident. The following section will make their meaning clear. § 9. Increase with reference to its Possible Outcome. The increase of divorce in this country has at length at- tracted attention and study. Its progress for the past twenty years has been marked out. During that time the current has been rising steadily and rapidly. But families living on a river's edge in spring are most anxious to know whether the stream is rising one inch an hour, or two, or six. It is their only means for determining how long a time will elapse before the torrent, unless its rise be stayed, shall sweep away their homes. Similarly the real importance of the increase of di- vorce lies in the indication it affords of the lapse of time before its rise will seriously threaten the stability of the particular homes in which each one is most interested. As marriages may end by two means, death and divorce, the simplest method of estimating the future growth of the lat- ter is by comparing the divorces with the total terminations of marriage. Mr. Wright states (p. 185) that the average dura- tion of married life in the United States, until ended by death, may be assumed to be about twenty-four years. Then the 20 THE DIVORCE PROBLEM. annual number of terminations of marriage by death may be considered one twenty-fourth of the number of couples. On this basis it is found that of the total of 314,350 terminations of marriage in this country in 1870, 96^ per cent, were by death and 3^ per cent, by divorce. In 1880 the percentage of divorces was 4.8, and in 1890 probably about 6.2. On the assumption that the conditions of the past twenty years remain unchanged, and the population continues to increase through the next century at the rate of 28 per cent, a decade, and divorces at the rate of 69.2 per cent., the latter will consti- tute a rapidly increasing percentage of the total terminations . of marriage, as follows : 1 900, 8 per cent. ; 1 9 1 o, 1 0.4 per cent. ; 1920, 13.3 per cent; 1930, 16.8 per cent; 1940, 21 per cent; 1950, 26.1 per cent; 1960,31.8 per cent; 1970, 38.2 per cent; 1980, 44.9 per cent; 1990, 52.1 per cent; 2000, 58.8 per cent At the end of one hundred years of increase like the last twenty, more marriages would end by divorce than by death. It is obvious that this would involve a fundamental alteration in the nature of the institution. It may be urged that the con- ditions will not be constant, that in fact they are rapidly alter- ing, and it must be admitted that the population will not con- tinue to increase at the rate of 28 per cent, in each decade. But thus far the progress of divorce has suffered no check. On the contrary its rate has slightly accelerated. The second quinquennial increased over the first 27.9 per cent. ; the third over the second 30.3 per cent; the fourth over the third 31.4 per cent. It is not claimed that such a computation has much value as determining the future. The elements are too varying and ill- determined to give good basis for a prediction. Those theo- retical considerations, stated in §§35, 36, must modify the results here stated, and that to an unknown degree. Such a calculation, however, has its importance as magnifying some ten diameters the past increase and making the meaning of its rate more obvious. THE DIVORCE PROBLEM. 21 § 10. Increase among the Negroes. The court records in the southern states seldom indicate the color of the parties to a suit and, therefore, no direct statistical evidence on the subject of negro divorces is accessible. Some study of the subject having led me to the belief that the in- crease among the negroes has been abnormally rapid, the following method of testing the hypothesis was adopted. It is essentially a comparison of the black with the white counties in the same state. If the increase among the negroes has been excessive, the counties containing a large percentage of that race would probably show on the whole an unusual in- crease of divorce. The counties of seven fairly typical south- ern states have been grouped in classes; the first including all counties with i to 7 per cent, negroes, the second all with- 7 to 17 per cent., the third all with 17 to 35 per cent, the fourth all with 35 to 60 per cent, and the fifth all with over 60 per cent. This grouping is based on the maps in the census of 1880. The divorce rate relative to married couples has then been determined for each group for 1870 and 1880, and the percentage of increase ascertained. The final result is given in the following table. The figures preceded by a minus sign indicate the percentage of decrease for the decade, the other figures give the percentage of increase. In several in- stances the first number against each state is based on so small a number of counties as to be less trustworthy than the others. When weighing the meaning of the table, it should be remem- bered that the average decennial increase for the whole coun- try was 69.2 per cent. TABLE II. Comparison of Decennial Increase in White and Black Counties. Counties with Counties with Counties with Counties with Counties with States. 1-7 per cent. 7-17 per cent. 17-35 per 35-60 per 60-100 per negroes. negroes. cent, negroes, cent, negroes, cent, negroes. Virginia — II .44 81 104 100 North Carolina. 28 27 78 114 146 Georgia — 34 10 30 16 43 Florida 10 34 104 88 164 Alabama 17 170 118 298 Mississippi .- 162 52 80 174 Arkansas 52 21 23 90 142 22 THE DIVORCE PROBLEM. In all these states but Georgia the counties containing over 35 per cent, of negroes show an increase greater, and in most instances much greater, than the average for the whole coun- try. On the other hand, in all the states but Mississippi (and there the first figure is untrustworthy because based on only 4 counties, with a total population of about 30,000), the coun- ties with less than 17 per cent, negroes show an increase less than the average for the country. The increase of divorce, be it remembered, is totally different from its absolute amount We shall have to examine later the question whether the negroes as a race are more addicted to divorce than southern whites. § 1 1. Increase in the Several States. The method employed in the last section for determining the increase among the negroes is here applied to the states. The number of divorces to 100,000 couples has been calculated for each state in 1870 and 1880,* and the percentage of in- crease for the decade ascertained. TABLE III. Decennial Increase in the Several States. per cent. Connecticut —26 Washington — 16 Idaho — 10 Kansas New York Nevada Vermont Rhode Island . . Delaware Indiana Maryland Wisconsin . . . . Pennsylvania 16 Oregon 17 Iowa 20 —9 —6 — 1 o 1 5 6 7 8 per cent. Wyoming 20 Massachusetts 21 Maine 22 Ohio 27 Kentucky 27 Illinois 30 Minnesota 30 United States 31 Georgia 33 West Virginia 37 Nebraska 39 Missouri . . 41 District of Columbia. 44 New Jersey 45 Michigan 50 per cent. Tennessee 52 Montana 53 New Hampshire 57 California 59 South Carolina .... 67 Virginia 73 North Carolina .... 100 Colorado in Louisiana 115 Arkansas 121 Dakota 123 Florida 131 Texas 139 Mississippi 142 Alabama 155 These percentages do not indicate the absolute increase, which in most cases is much greater, but the gain on the pop- ulation, a far different and more significant matter. All the * Compare Table IX, § 20. THE DIVORCE PROBLEM. 23 southern states containing many negroes, except the three border states of Delaware, Maryland and Kentucky, have in- creased more rapidly than the average. The rapidity of in- crease culminates in Florida and the states about the lower Mississippi : — Alabama, Mississippi, Arkansas, Louisiana and Texas. On the other hand, nearly all the states north of Mason and Dixon's line and east of the Mississippi have in- creased less than the country as a whole. The exceptions are New Hampshire, New Jersey and Michigan. The states west of the Mississippi are interspersed irregularly in the table. Therefore divorce is increasing most rapidly in the southwest and south, least rapidly in the northeast and north, and with irregular rapidity in the trans-Mississippi states. The real increase in the south has been even greater than indicated by the table. The errors of the census of 1870 made the population of many southern states too small. Ob- viously, a divorce rate calculated upon that population would be too large ; the increase over it shown by the true rate for 1880 too small. § 12. Duration of Marriage before Divorce. In 92.7 per cent, of the cases the date of the ceremony and, accordingly, the duration of the marriage, are known. The average length of these 304,726 marriages ending in divorce is given by Mr. Wright as 9.17 years (p. 183). It does not follow, however, as might at first be supposed, that one-half of them continued that length of time before divorce — not at all ! In one-half the cases the divorce came after only 6.56 years of married life, but the other half varied in length from 6 to 50 years and so raised the average to 9.17. For example, ima- gine a county in which nine divorces were granted, eight at the end of three years of married life, and the ninth thirty years after marriage. Obviously, the average duration of the nine would be six years, and yet it would be of greater socio- logical interest and significance to know that eight-ninths were granted at the end of three years. One-fourth of all the di- 24 THE DIVORCE PROBLEM. vorces came within three and one-half (3.42) years after mar- riage ; one-third within about four and one-third (4.36) years ; two-thirds within nine and two-thirds (9.66) years ; and three- fourths within eleven and five-sixths (11.83) years. The lapse of time between the_separation of the parties and the divorce was ascertained in 45 selected counties. In half these cases it was less than 1.7 years. Assuming that the results for the whole country would agree with this, it follows that in half the instances the separation of the parties took place within 4.86 years after marriage. This confirms, more than at first sight Mr. Wright's figures seem to do, the popular opinion that estrangement and separation are most frequent in the early years of married life. The average length of marriage before divorce varies widely in the different parts of the country. This is shown by the fol- lowing table : TABLE IV. Average Length of Mafriage before Divorce. years. Arkansas 6.4.8 Tennessee 6.91 Louisiana 7.57 Kentucky 7.59 Mississippi . . . 7.75 Missouri 7.78 Indiana 7.78 Texas 8.07 Florida 8.13 Kansas 8.19 Alabama 8.32 Wyoming. 8.45 Nebraska 8.47 Illinois 8.59 Georgia 8.67 years. Montana 8.83 Idaho 8.93 Virginia 8.96 Nevada 8.97 Colorado 9.03 United States 9.17 West Virginia 9.37 Iowa 9.46 Ohio 9.60 Michigan 9.62 Wisconsin 9.82 California 9.83 Washington 9.90 North Carolina.. . . 9.93 Maryland 9.94 years. Rhode Island 9.97 Pennsylvania 9.98 Connecticut 10.04 Oregon 10.07 New York 10.21 Dist. of Columbia . 10.29 Maine 10.29 Delaware 1040 Dakota 10.54 New Hampshire . . 10.69 Vermont 10.80 Minnesota 10.84 New Jersey 11.69 ! Massachusetts 12.12 The interval between marriage and divorce is shortest in the southern states, and longest in the northern states east of the Mississippi. It is probable that the interval in Europe is somewhat longer than the average for our whole country. Ten years of Swiss statistics indicate an average of 9.8 years * and M. Bertillon (p. 20) says the average is about ten years. * Die Bewegung der Bevolkerung in der Schweiz im Jahre 1885. Beilage I. THE DIVORCE PROBLEM. 25 Differences of legislation are influential upon this element. For example, Massachusetts continued much longer than her neighbors to require that desertion should last for five years in order to become a ground of divorce. Consequently, the average interval in divorces of this class, constituting 45 per cent, of the whole number, was a full year longer in Massa- chusetts than in the adjacent New England states. It is encouraging to find that the duration of marriage be- fore divorce is increasing with considerable rapidity, both in the country as a whole and, so far as my examination has gone, in each of the states. In the first five years the average was 8.86 years ; in the last five, 9.59; in the first, one-half the divorces were issued within 6. 1 3 years after marriage ; in the last, one-half within 6.83. The average has thus increased nearly nine months in fifteen years. The following table will indicate the facts more clearly : TABLE V. Increase of Divorces Grouped According to the Duration of Marriage. Duration of marriage in years. Divorces granted, 1867-71. Divorces granted, 1882-86. Per cent, of increase. 2899 4019 4974 4746 4068 3388 3«>5 2516 2133 2051 1778 1 539 1398 1 189 "34 934 914 742 694 645 4038 53H 7483 9426 9671 9014 8274 7021 6093 S305 5002 4384 4089 3563 3«44 2931 2721 2217 1877 1577 1459 9401 83 86 89 104 122 £:::::::::: 144 134 142 149 8 T/ M 146 15° 164 158 16 191 143 18 153 127 126 21 and over.. «33 49,004 109,966 119 2 6 THE DIVORCE PROBLEM. It thus appears that the average increase for the fifteen years was 1 19 per cent., but the increase of divorces granted within four years after marriage was considerably less than this. On the other hand, those granted more than four years after mar- riage show a larger increase, and the climax is reached in those granted after between twelve and eighteen years of married life. These increased between 143 and 191 per cent. § 13. Remarriage after Divorce. Over and over in the progress of the discussion the asser- tion has been made that most divorces are obtained in order to open the way to a second marriage, and the conclusion has been drawn that a limitation on remarriage, or a denial of it altogether, would greatly reduce the number of divorces. A most emphatic statement of this character was made in a recent article by Hon. E. J. Phelps. He says : * " The question is not whether divorce laws shall exist, but whether they shall permit the divorced parties to remarry. If that right were taken away, nine-tenths, perhaps ninety-nine hundredths, of the divorce cases would at once disappear. In the vast majority of in- stances the desire on the part of one or the other or both to remarry is the foundation of the whole proceeding." No proof is offered of the dogmatic assertion. Mr. Phelps says it is " the result of a long observation of judicial proceed- ings in this class of cases;" but the observation of any single lawyer, however eminent, is so inadequate a basis for so broad a generalization, that it deserves only to be laughed out of court. His opinion, however, is shared by many excellent people and merits serious attention. No clue to an answer can be found on this side of the water. Having no good statistics of marriage, we have no statistics of remarriage. But the con- current testimony of the statistics of Berlin, Holland and Switzerland replies to the question. Is it asked: How shall it be investigated by the statistical method? The thing is very simple. Marriages end by death or divorce. Now, it is * Forum, Dec, 1889, p. 352. THE DIVORCE PROBLEM. 2 7 presumable that the death of a husband or wife does not occur, in any appreciable number of cases, as a result of the surviving partner's desire to marry again. Therefore, the interval between the death of a husband or wife and the re- marriage of a widow or widower may be considered the normal interval, and any diminution of this would register the influence on divorce exerted by a present desire of remarriage. The point has been examined in detail by M. Bertillon in an article, " Du sort des divorces" originally published in June, 1884, in the Journal de la Societe de Statistique de Paris, and translated the same year for the Journal of the Statistical Society of London. As the article in either form is somewhat inaccessible, one table is here reproduced. That of Switzer- land is selected, because the conditions and divorce rate of that country correspond most closely to our own. TABLE VI. Remarriage of Divorced Persons in Switzerland. Time between end of first and beginning of second marriage. Of 1000 widow- ers remarry- ing within ten years, number rema r r y i n g each year. Of 1000 such divorced men, number re- marrying each year. Of # 1000 such widows, num- ber remarry- ing each year. Of 1000 such di- vorced women, number re- marrying each year. Less than 1 year 323 260 136 82 48 108 43 300 255 'Si 106 S3 101 34 95 264 152 132 9' 196 70 194 282 166 127 68 125 38 1000 1000 1000 Could it be proved more conclusively that divorced men and women are not more disposed to marry directly after the de- cree than widows and widowers are tS marry after the death of husband or wife. Another series of tables has been compiled by the same 2 8 THE DIVORCE PROBLEM. statistician, comparing the tendency to marry shown by bachelors, widowers and divorced men, and by spinsters, widows and divorced women. Until the age of 35. he finds, divorced men and women marry less often proportionately than widowers and widows, but after that age they show a greater tendency to marry. For a detailed proof of this posi- tion, the reader is referred to the article already cited. § 14. Distribution of Divorce between Natives and Foreign Born. The censuses of Massachusetts and New York divide the residents of those states into the three classes of single, married and divorced, and subdivide them into natives and foreign born. Obviously, divorced persons who have remarried would be included in the second class; many others would return themselves as single in preference to divorced. Therefore, these data are very untrustworthy ; but, on the other hand, the figures in the two states agree in indicating the ratio of divorced natives to married natives to be about four times that of the divorced foreign born to the married foreign born. In the absence of other and more accurate data, these maybe used to give a rough approximation. The New York census indi- cates also the marriage rate of the foreign born. Sixty per cent, are married, 52 per cent, to a foreign born husband or wife, 8 per cent, to a native. In default of further information this must also be extended to the whole country. On that as- sumption the 6,679,943 foreign born residents of the United States in 1880, would be married as follows: 3,473,570 among themselves and 534,395 to natives, while the remaining 2,671,- 978 would be unmarried. When compared with Mr. Wright's estimate of the total number of couples in the country, this gives 7,193,728 native couples, 534,395 mixed couples and I .736,785 foreign born couples. It must be assumed that mar- riages of mixed couples are as likely to be broken by divorce as those of native couples. The probability is that they are somewhat more so, but no measure of the excess is obtainable. Then the 19,228 divorces of 1880 would be distributed as fol- THE DIVORCE PROBLEM. 2 Q lows: 1,023 to foreign born, 1,259 to mixed and 16,945 to native couples. While the ratio of divorces to 100,000 couples for the whole country was 203, that ratio for the native couples alone was 236. § 15. Distribution between Catholics and non-Catholics. Our Catholic population is an element of the problem about which even less is statistically determined than about the for- eign born residents. Even the number of Catholics in the country has never been given in any national census, and, therefore, it is necessary to accept the figures of the Catholic Directory, 6,832,954. Because of the lack of information, the important and complicating element of mixed marriages must be disregarded ; and the assumption made that there were in this country in 1880 about 1,291,428 Catholic couples, and 8,173,480 non-Catholic (we can hardly say in all cases Pro- testant) couples. The carefully collected statistics of Switzer- land show 73 divorces a year to every 100,000 Catholic couples, 283 a year to every 100,000 Protestant couples and 605 a year to every 100,000 couples of mixed religion (Bertil- lon, p. 32). In default of statistics for our own country, noth- ing better can be done than to apply these ratios and assume that in this country also divorce is four times as frequent among Protestants as Catholics. The rough approximation thus obtained is as follows : in 1880 there were 18,497 divorces among the 8,173,480 non-Catholic couples, and only 731 among the 1,291,428 Catholic couples. It is probable that the real difference was even greater than this, because the dif- ference of religion would be accentuated in many cases by a difference of race and social position. § 16. Distribution between Negroes and Whiles. The common opinion in the southern states is that divorce there is almost confined to the negroes. Mr. Wright gives expression and qualified indorsement to this idea as follows : , THE DIVORCE PROBLEM. "It is probably true that in nearly all the (12) states named, where the colored population is very dense, nearly, if not quite, three-fourths of the divorces granted were to colored people ; at least this is the evidence furnished the Department by clerks of courts and by others in a position to judge with fair accuracy." (p. 132.) The states named have from 16 to 61 per cent, negroes. As already stated, the records seldom indicate the color of the parties. An a priori argument against this opinion may be derived from what is known of divorce in other parts of the world. It is not the poorest and most ignorant classes that frequent the divorce courts, their poverty and ignorance prevent. Among them a change of husbands and wives occurs, if it occurs at all, without appeal to the law. This general fact would apply with especial force to the negroes, because they have been trained into licentiousness by slavery. In the old days, the slave's marriage was governed by his master's whim, and the first effect of freedom would natur- ally be to substitute his own whim for that of an owner. Is it not strange, even inexplicable, that the negroes, three or four years after emancipation, began flocking to the courts to have their marriages legalized? Against statistics such an argument would be useless, but against the observation even of clerks of courts it is entitled to some weight. For such observation would be almost valueless except as to the present, while Mr. Wright asks us to believe that for the whole score of years three-fourths of the divorces have been granted to negroes. Perhaps the statistics may be so turned about as to furnish indirect evidence, in place of the direct testimony they with- hold. If the contention that the negroes monopolize the di- vorces be true, wherever the former are most numerous the latter would probably abound. The states with the largest per- centage of negroes would have the greatest amount of divorce. But there is no relation between these two phenomena. Here are the twelve states mentioned by Mr. Wright, arranged in the order of their percentage of negro population, and against each is placed its rate of divorce as explained in § 20. THE DIVORCE PROBLEM. Per cent. Divorce of Negroes, rate. South Carolina 61 Mississippi 58 Louisiana 51 Alabama 48 Ilorida 47 Georgia 47 Virginia 42 . North Carolina 38 . Arkansas 26. Tennessee 26. Texas 25 . Kentucky 16. 31 Per cent. Divorce of Negroes, rate. 58 3° 270 191 263 188 • s • 157 • 75 ■ H3 . 287 • 77 Were any conclusions at all to be drawn from this, it would be that those states, like Arkansas, Tennessee, Texas and Kentucky, with fewest negroes, have most divorces. But it is objected, no conclusion of any kind can be drawn. The number of divorces is controlled by the legislation of each state, and of course South Carolina heads the list. It will be shown in the second part of the discussion that the objection is true only in part. For the present, however, its cogency may be admitted, and the argument restated in such form as to avoid the difficulty. Since no differences of legislation exist between different counties of the same state, a comparison be- tween counties is not open to this objection. In seven states the counties have been grouped as already explained in § 10, and the annual number of divorces to 100,000 couples com- puted for each group for both decades. The results are stated in the following table : TABLE VII. Divorce in White and Black Counties. States. Counties with Counties with Counties with Counties with Counties with 1-7 per cent, negroes. 7-17 per cent, negroes. 17-35 per cent, negroes. 35-60 per cent, negroes. over 60 per cent, negroes. Virginia North Carolina . Georgia Florida Alabama Arkansas . 149 67 338 First Decade, 1867-7(1. *r '51 68 33 92 208 104 4i in Virginia North Carolina . Georgia Florida Alabama Mississippi Arkansas Second Decade, iSff-Sb. 36 18 74 168 116 231 27 H 79 191 73 58 202 132 98' 6 S ss 86 42 32 30 103 101 96 92 37« 379 342 360 122 208 159 108 177 104 231 135 282 384 24 13 44 114 45 35 186 48 32 61 301 179 94 45' , 2 THE DIVORCE PROBLEM. From the differences between these two sets of figures the percentage of increase in the white and black counties already given as Table II (§ 10) was calculated. The present table in- dicates that in all the states but Arkansas the divorce rate was less in the black counties than in the white. It may be inferred with reasonable confidence that the negroes do not obtain divorces with so much greater frequency than the whites as is commonly supposed. It may be objected, the social and economic conditions of the diverse counties vary so widely that the result is entirely un- trustworthy. The anwer is, that where so many counties are included in each group (in most cases from 15 to 40 with a population of several hundred thousand), the other differences are in large measure eliminated; and no reasonable cause for the uniformity of result can be found, unless it be the uniform variation of the single element that does vary uniformly, namely, the percentage of negroes. The only instance in which the negro element is almost entirely absent is found in two small counties of Florida, Polk and Manatee, each with about 4 per cent, of colored citizens. They have about 1,160 white couples and 51 negro couples, and are charged with '53 divorces for the twenty years. Very few of these can have gone to the blacks, and yet the rate for the two counties, 267, is almost identical with the average rate in Florida, 287. On the whole, it seems probable that the average negro rate is rather below that of the southern whites, but is increasing much more rapidly than the other, and in a few localities or states may have already reached or passed it. § 17. Distribution Between City and Country. Mr. Wright proves that divorce is more common in cities than in the country, but the method he was compelled to fol- low is so rude that it fails to fix the amount of difference. In fact, many rural counties have a rate as high as, or higher than, that of the largest city of the state. For example, Berkshire and Hampden counties, Massachusetts, have a rate THE DIVORCE PROBLEM. ,3 considerably above that of Suffolk; and four other counties in that state have a rate practically identical with the Boston one. Now the European statistics prove conclusively that the city rate is from three to five times that of the surrounding country (Bertillon, page 5 5). To reconcile the discrepancy between these figures and ours, it must be assumed that some counter- balancing influences are at work here. Our foreign born and Catholic population have a divorce rate about one-fourth that of the rest of the country, and they are so massed in the cities as appreciably to affect the city rate. The theory that their presence masks the true influence of city life derives support from the figures of the state censuses; but those data are to©, inaccurate to be cited in detail. § 1 8. Distribution between Couples with Children and those without Children. It is probable that the presence of children in a family di- minishes the tendency to divorce. Can statistics test this theory and give any measure of the influence thus exerted ? If the number of childless couples, the number with children and also the number of divorces in each class, were known, then the tendency to divorce in each might be calculated, and the difference, if any, be found. A roug*h approximation to- an answer may be obtained by combining some figures in the- Report with others in the Massachusetts census of 1885.. That state is, perhaps, the only one to report the number of childless wives and wives with children. The former are 17.56 per cent, of the whole number of married women. There is only one state, New Jersey, in which the court records indi- cate with fair completeness whether the parties divorced had or had not children. In that state the number reported as un- known was only 2.6 per cent. Three other states, Minnesota, Maryland and Colorado, report as unknown, respectively, 8 per cent, 14 per cent., 14 per cent. All the rest have records so incomplete as to be worthless for our purpose. By a com- bination of these figures with the estimated number of couples 3 45-8 I6S-S 3-6 96.9 369.8 3-« 4°-3 140.5 3-5 260.7 1 42 1. 6 5-4 34 THE DIVORCE PROBLEM. for 1875 and the Massachusetts ratio of childless wives to the whole, the following results are obtained : Divorces annually Divorces annually to 100,000 cou-i to 100,000 child- jjario pies with chil- less couples. dren. | New Jersey . Minnesota . ■ Maryland . . Colorado . . . The errors involved in applying to a frontier state like Colorado a Massachusetts birth rate and an inaccurate ratio of couples to population, determined from a few eastern states, may account for the variation shown. On fop wT 1 " 1 ^ |T ■'" f air to conclude, notwithstanding, that childless marriages ar e be- t ween three and four times as likely to end in divorce as" marriages with children. The only foreign census citeaBy M. Hertillon (p. 135), as giving the same datum as that of Massachusetts, is the French census of 1856. The results obtained from that confirm the foregoing. The rate at which separations were de manded was 3.6 times as gTcal ainuhg c hildless couples as among others. § 19. Distribution of Divorces as granted to Husband or Wife. Nearly two-thirds of the divorces (65.8 per cent.) were granted on demand of the wife. The party going into court is usually somewhat more innocent than the other. Therefore- the husband more often destroys the marriage tie than the wife ; certainly he is more likely to commit those overt acts on which a suit for divorce must usually be based. Law is for the protection of the weak against the strong through the superior strength of numbers, i. e. the community. But weakness may be beyond the effective protection of law. A modicum of knowledge and courage is necessary, or law is of no avail. Seldom can the police effectively restrain parental THE DIVORCE PROBLEM. 35 abuse, and the submissiveness of negroes has foiled many an effort at interference on their behalf. In some parts of the country and some classes of society, a wife's relation to a hus- band is that of a child to a parent, or a servant to a master, much more than of an equal to an equal. Public opinion, her early training, dread of notoriety and the absence of all means of support apart from her husband, compel her to endure wrongs and ill-treatment to which the ear of the judge would be open. The purpose of the divorce law is foiled. Hence the number of divorces decreed is no more a test of the wrongs endured by wives, than the number of parents pun- ished for cruelty to children is a criterion of the abuse they suffer. Divorces to wives measure their resistance, not their burdens. Southern wives are probably more resigned and submissive than northern, and, in accordance with this fact, the Report shows that they are less inclined to seek a divorce. In seven southern states, less than half the divorces are granted to the wife, and in the whole sixteen the ratio is only 55.4, while in thirty northern and western states and terri- tories 69.2 per cent, are granted to women. If the number of divorces granted to the husband had been unchanged, 112,540, but those granted to wives had remained through the twenty years at the southern ratio, 55.4 percent, of the whole, the total number would have been 251,776, instead of 318,716, and the number granted to northern women 95,240, instead of 172,183. The differences between the various states and territories in this regard may best be indicated by the following table: 36 THE DIVORCE PROBLEM. TABLE VIII. Percentage of Divorces to Wife in Stales and Territories. North Carolina 39.3 Mississippi 40.0 Alabama 42.5 Virginia 43.9 South Carolina 45.4 Florida 45.8 West Virginia 48.0 Georgia 51.8 Arkansas 52.4 Texas 554 Utah* S54 Louisiana 56.2 Dakota 57.2 Kentucky 58.8 Maryland 604 Tennessee 61.2 Delaware 62.3 New York 62.6 New Jersey 62.9 Missouri 64.6 New Mexico 65.1 United States 65.8 New Hampshire .... 65.9 Pennsylvania 66.5 Arizona 66.6 Colorado 67.0 Wisconsin 67.4 Nebraska 68.3 Iowa 68.4 Minnesota 68.5 Wyoming 68.8 Illinois 68.8 Vermont 69.2 Kansas 69.3 Massachusetts 69.5 Michigan 69.9 Connecticut 70.6 Maine 70.7 Indiana 71.0 Washington 71. 1 District of Columbia . 72.1 Ohio 72.2 Oregon 72.5 Idaho 72.6 Utah * 74.1 Montana 74.8 California 75.3 Rhode Island 77.6 Nevada 77.9 The average length of marriages ending in a divorce granted to the wife is somewhat greater than of those where the divorce is granted to the husband. In other words, wives show a little more patience and long-suffering than husbands. The differ- ence, however, is but trifling, three-tenths of a year, and even this is not found all over the country. Of the eighteen states and territories west of the Mississippi, all but Missouri, Ne- vada, and Utah have a shorter average interval between mar- riage and divorce in cases where the wife is the complainant. On the other hand, of the twenty-seven states east of that river, including Louisiana, Minnesota and the District of Columbia, twenty-one show a longer duration of marriage when divorce is decreed to the wife. The six exceptions are three states in the southeast, Georgia, Florida and Alabama, two in the northwest, Michigan and Wisconsin, and one in the northeast, Rhode Island. The Mississippi, then, divides the country into two belts, an eastern, where suffering wives show a little more ♦From 1875 to 1879 hundreds of divorces for non-resident eastern parties were granted in Utah (compare § 28). Most of these, fully seven-tenths, were to the husband. Hence the true position of Utah is found by neglecting the figures for those four years, and computing the percentage from the other sixteen. THE DIVORCE PROBLEM. . 37 patience than husbands, and a western, where the reverse is true. § 20. Distribution among the Several States. The divorce rates of the states given in the table below have been computed as follows. The number of divorces in 1870 and 1 880 was determined by averaging the figures for the seven years, 1867-73 and 1877-83 ; this average was divided in each case by the estimated number of married couples in the state for that year, and the decimal thus obtained was mul- tiplied by 100,000. The result obviously expresses the num- ber of divorces annually to 100,000 couples. TABLE IX. Divorce Rate of the States in 1870 and 1880. South Carolina 3 North Carolina 15 Louisiana 26 Virginia 33 New Jersey 47 Alabama 56 Delaware 57 Georgia 58 Maryland 60 Mississippi 65 New York 86 West Virginia 95 Pennsylvania 96 1870. Minnesota ill Texas 113 District of Columbia. 123 Florida 1 24 Arkansas 130 Dakota 132 Tennessee 133 Massachusetts 133 Kentucky 148 Missouri 153 United States 155 Nebraska 169 Ohio 196 Wisconsin 202 Michigan 253 Kansas 257 Vermont 263 Iowa 264 Illinois 274 New Hampshire 280 California 289 Maine 326 Indiana 349 Colorado 371 Idaho 376 Montana 448 Connecticut 453 Oregon , 460 Rhode Island 470 Washington 525 Wyoming 528 Nevada 611 38 . THE DIVORCE PROBLEM. South Carolina 5 North Carolina 30 Virginia 58 Delaware 60 Maryland. • • • ■ 64 New Jersey 68 Louisiana 75 Georgia 77 New York 81 Pennsylvania Ill West Virginia 130 1880. Alabama 143 Minnesota 145 Mississippi 157 Massachusetts. 161 Dist. of Columbia . . 177 Kentucky 188 Tennessee 191 United States 203 Missouri 216 Wisconsin 218 Kansas 235 Nebraska 235 Ohio 249 Vermont 260 Texas 263 Arkansas 270 Florida 287 Dakota 294 Iowa 316 Connecticut 333 Illinois 357 Indiana 369 Michigan 379 Maine 399 Idaho 433 New Hampshire . . . 439 Washington 441 California 459 Rhode Island 476 Oregon 538 Nevada 606 Wyoming 633 Montana 685 Colorado 781 The returns foom Louisiana and Idaho are incomplete. No report was received from counties containing about one-fourth the population of the former and one-fifth that of the latter. In the table for 1880 a correction has been made by assuming that the rate in these counties was the same as the average for the rest of the state. The breaking up of each table into three irregular columns has a purpose. It aims to attract attention to a most sugges- tive conclusion indicated by the table. Divorce is distributed over the country along geographical lines. The division is not made by parallels of latitude into north and south. The lines of cleavage in this case are the two great mountain systems of the country dividing the states into Atlantic, Pacific and cen- tral. Divorce is most common on the Pacific, least common on the Atlantic. Study the table a moment, and note how remarkable a division it is. Observe that the first column in- cludes the seventeen Atlantic states, except Florida and the six in New England ; the second, the eighteen central states, except Louisiana ; the third, the eight Rocky Mountain and THE DIVORCE PROBLEM. ,g Pacific coast states. The great exception is found in the six states of New England. They show more divorce than would be expected from their geographical position. Compare the two tables, however, and it will be seen that ever}' one of them but New Hampshire has ascended in the scale and drawn nearer the first group, to which it naturally belongs. How- ever, it must be admitted that they constitute an exception to the general division, which must find a deeper explanation in the chacteristics of New England blood. Look for a moment at the other deviations from the rule. The Catholic influence may contribute to explain the free- dom from divorce in Louisiana. The District of Columbia is really a city, about 83 per cent, of its population is urban, and it only confirms the general theory of the influence of urban life to find the divorce rate in the District nearly three times that of the adjacent states, Maryland and Virginia. Hence this exception is merely apparent. The Rhode Island rate, higher than that of any other state east of the Rocky Moun- tains, may find a partial explanation in a similar fact. That state has a proportion of urban population (jj per cent.) larger than any other. Since the war there has been a large migration to Florida from the north, which may have exerted an influence on the divorce rate there. A slight statistical indication that this conjecture is correct may be found by comparing the place of marriage of the divorced parties in Georgia, Florida and Alabama. Fifteen per cent, of the Florida divorces were granted to parties married elsewhere ; only 5 per cent, were granted to such parties in Georgia and Alabama, and nearly all of these latter came from adjacent states, only about 1 per cent, coming from a greater distance. In Florida, on the con- trary, about 1 1 per cent, were granted to parties married either in a non-adjacent state or in a foreign country. Neglecting the nine exceptions italicized in the table, the Atlantic coast state with highest divorce rate has less than the 40 THE DIVORCE PROBLEM. central state with least, and the central state with most has less than the western state with least. The division is so regular and clear that the question at once arises, what is the cause ? Is any explanation to be found in the different* systems of legislation in vogue in the three belts ? The query is but a phase of the larger problem: What is the influence of legislation on divorce ? In this form it will constitute the sub- ject of the second part. PART II. INFLUENCE OF LEGISLATION ON DIVORCE. § 21. Problem and Method. Law and other causes influence the divorce rate. Thus much is axiomatic. But when an attempt is made to go fur- ther and determine the relative influence and effect of law and the sum of other causes, then the controversy opens. One party claims that law is a minor and relatively unimportant factor; another, that it is the controlling element. Between these two views a decision^ must be made. That is the prob- lem. How is it to be solved? In two ways. First, select states or countries with similar social and economic condi- tions, but very different laws, and compare their divorce rate; do the same for states with similar laws, but different eco- nomic conditions ; note whether the divorce rate varies with the law, or with the other factors, or with neither exclusively. Secondly, examine every instance of a change in the divorce law, and observe whether it was attended by a change in the figures such as might have been produced by the law. To prove the point there must be a uniform conjunction of such changes, and it is constantly to be remembered that negative cases are far more weighty than positive. For the latter may be easily explained as mere coincidences, unless uniformity of occurrence makes any such explanation impossible. § 22. Legal Causes of Divorce. The forms of legal procedure in divorce cases are an un- inviting subject of study. A large measure of uniformity in substance underlies the differences in the various states, and (40 42 THE DIVORCE PROBLEM. there is no evidence to prove that the superficial variations have exerted a marked influence on the divorce rate. Con- sequently they will here be neglected. But with reference to the legal grounds of divorce, the matter stands otherwise. The number of these differs widely in the various states. Does the divorce rate show corresponding variations ? As a preliminary step, the grounds must be analyzed and grouped. Like the forms of procedure, they manifest super- ficial variations covering fundamental uniformity. Mr. Wright has reproduced the former, and not clearly shown the latter. He enumerates forty-two grounds of absolute divorce, each in force in some part of the country, but a single ele- mentary distinction will reduce the number materially. Divorce is the dissolution of legal marriage ; if there has been no mar- riage, there can be no divorce. A court may declare the pre- tended marriage a nullity, but such an adjudication is in no proper sense divorce, although the statutes in many of the states so term it. The misuse of English by state legisla- tures, however, must not be permitted to obscure the difference in things, and such a designation can be retained only with a protest expressed by styling it an improper divorce. The grounds on which improper divorces are granted include such causes as certain degrees of relationship, lack of age, capacity, or consent. In not a few of the states the list of such im- pediments to marriage laid down by the common law has been extended by statute. Rather more than one-third of the forty-two grounds of absolute divorce enumerated by Mr. Wright are really grounds invalidating the marriage from the beginning. All these and the improper divorces granted for them, 2,854 for the twenty years, must be disregarded. The remaining causes may be grouped as general or local ; each of the six former is admitted in at least frwrnfy fivf^" states, while each of the eight latter is confined to one or two. The number and per cent, of divorces for each cause appears in the following table : THE DIVORCE PROBLEM. 43 TABLE X. Divorces Classified by Causes. Total Divorces 328,716 Cause Unknown lo,3 I S Improper Divorces 2,854 13,169 Proper Divorces for Known Cause 315,547 General Causes. Desertion 126,676 Adultery 67,686 Cruelty 51,595 Drunkenness 13,866 Neglect to Provide 7,955 Imprisonment! 2,721 Combination of General Causes 35,417 Total for General Causes 305,916 Local Causes. Neglect of Duty, Ohio 2,685 Incompatibility of Temper, Utah 1,579 Misconduct, Connecticut Voluntary Separation, Wisconsin Violent Temper, Florida Vagrancy, Missouri Supervenient Insanity, Ark., Wash Gross Misbehavior and Wickedness, Rhode Island 454 228 119 28 Total for Local Causes 5, 149 Combinations of General and Local Causes 4,1 15 Minor Causes 367 3IS.S47 Per cent. 40.15 21-45 16-35 4.40 2.52 .87 11.23 96.97 1.60 1-3' The local causes may be dismissed in few words. Two have been repealed in the course of the twenty years, incompatibility of temper in Utah and misconduct in Connecticut. Neglect of duty in Ohio is substantially the same as neglect to provide in other states, but apparently is construed somewhat more broadly, for husbands sue on the ground of neglect of duty, while only wives have a legal claim to be provided for. On the other hand, the Kansas statute couched in the same terms is apparently construed there as synonymous with neglect to provide, and divorces are reported for the latter rather than the former cause. It is probable that most of the 44 THE DIVORCE PROBLEM. 4,115 divorces for a combination of general and local causes, were really granted for the former. While Connecticut, for ex- ample, reported 1,107 divorces for desertion and misconduct together, desertion must have been the effective cause of the great majority. Therefore, about 98 per cent, of all divorces for known causes may be set to the account of the six general grounds, and to these our attention may be confined. Each of these grounds is wide-spread, but none universally admitted, for South Carolina has no divorce law. There are forty-eight jurisdictions and systems of law in our country, exclusive of Alaska. Forty-seven allow divorce for adultery, forty-five for desertion, forty-one for cruelty, thirty-nine for imprisonment, thirty-seven for habitual drunkenness, twenty- seven (including Kansas) for neglect to provide. The question now recurs, Does the distribution of causes of divorce throw any light on the geographical division stated in §20? § 23. Comparison of Number of Causes with Divorce Rate. The six general causes are distributed with considerable uniformity over the central and western belts. No explana- tion of the different rates in these two belts can be found in any difference of causes. The only one in reference to which there is much difference, is neglect to provide. This is ad- mitted by some and rejected by others, but no variation of rate can be shown to result, nor is the tendency to group along geographical lines thereby affected. Illinois and Iowa admit this cause, Indiana and Michigan do not, yet the former just outrank the latter. Tennessee admits it, Kentucky not, but there is only an inappreciable difference of rate. Kansas ad- mits it, Nebraska not, yet their rates are exactly identical. Comparatively little evidence is to be found in these two groups, but what there is goes to show that the differences of law are overridden by similar social and economic conditions. For in a large and general way it must be admitted that in such respects adjacent states are more similar than far distant ones. THE DIVORCE PROBLEM, 45 The eastern belt of states, from New York to Georgia, includes all which have manifested a conservative spirit in their divorce legislation. Here are found South Carolina with no divorce, North Carolina and New York with only one gen- eral cause, New Jersey and Maryland with two, Virginia and West Virginia with three. Apparently, then, these states show clearly the marked influence of legislation. But such a conclusion must not be too hastily drawn. These large and old states are very conservative and averse to radical changes of law. Their social and economic conditions have probably changed less than those of the other two groups. Certainly this is true of the southern half of the belt, where divorce is least frequent. So a closer examination is necessary. Let the states be arranged as below in the order of their freedom from divorce, and against each be placed the number of general causes for divorce allowed, and it will appear that there is no connection between the two. General causes States. admitted. North Carolina 1 Virginia 3 Delaware 6 Maryland 2 New Jersey 2 Louisiana 5 Georgia 5 New York I Pennsylvania ■ 4 West Virginia 3 New York, New Jersey and Pennsylvania, may be com- pared more closely. The New York rate is greater than New Jersey's, and only just above Pennsylvania's. This means that more divorces for adultery are granted in New York, relatively to population, than for adultery and desertion in New Jersey, and almost as many as for adultery, desertion, cruelty and im- prisonment in Pennsylvania. Assume the number of married couples in the three states in 1875 to be a mean between the 46 THE DIVORCE PROBLEM. estimates for 1870 and 1 880, and compare with this mean the total number of divorces for adultery in the three states for the twenty years. Pennsylvania had annually' 1 5-8 such divorces to ioo.ooo couples, New Jersey had 256, and New York, 781. Judging, from the court records one would say that adultery was about three times as frequent in New York as in New Jersey, and about five times as frequent as in Pennsylvania. No such inference is warranted. The true conclusion is that lim- iting the causes, increases the number of divorces in those which remain, but without materially affecting the total num- ber. A certain proportion of the married couples in the three states desired divorce and were willing to offer the evi- dence required in order to obtain the decree. The number of causes, then, seems to have affected the distribution of divor- ces, but in no large degree the total number. The law of Delaware is different from that of any other state. Here alone has survived the English custom prior to 1858 of granting divorces by Act of Parliament. In most of the states and territories such private bills are forbidden, and in those where it has not been prohibited the custom has fallen into disuse. But in Delaware about 80 per cent, of the di- vorces are granted by the legislature, and hence in theory may be for any cause. The largest percentage of cases re- ported as for cause unknown comes from this state. The custom appears to be lax. Moreover, in addition to this appar- ent looseness, the Delaware courts are empowered to grant di- vorces for any one of the six general causes. Even neglect to provide there is recognized as a ground, although not admitted in any other state of the eastern belt. We should expect accordingly to find divorce frequent in Delaware. But the opposite is true. The state ranks above its immediate neighbors, Maryland and New Jersey, and but very slightly below Virginia, while it has only about half the rate of Penn- sylvania. On comparing the number of divorces for adultery in Delaware with that for the neighboring states, we find New THE DIVORCE PROBLEM. d7 ^\ U 1ST* lb.<\' Jersey has 1,256; Maryland, 253, but Delaware, only 109 to 100,000 couples. The reduction of the number of causes from six to two seems to have had little influence on the total di- vorce rate, but to have made those granted for adultery about two and one-half times as numerous. The low rate of North Carolina appears at first glance con- nected with her one ground of divorce as effect with cause, yet even here the relation is doubtful. It may well be that she shares in the sentiment of her sister state to the south, that her public opinion likewise vigorously opposes divorce in any case, and that her rate is controlled by this opinion much more than by law. The three belts of states do not in any way conform to the differences in the legal grounds of divorce. Such differences are uniformly overridden. The clear tendency of states in geographical proximity to stand in juxtaposition in the table is equally unexplained by any similarity in their divorce law. 'Therefore law must be a relatively unimportant factor in the complete explanation of the variations of divorce rate in the states. § 24. Changes of Law in the Last Score of Years. The second method of examining statistically the influence of law on divorce is by a comparison of every change in divorce law with the figures. If a corresponding change uniformly appears, it may fairly be regarded as a direct effect. The ex- amination must clearly be limited to the legal changes in the twenty years for which statistics are given. Mr. Wright dis- cusses the question at some length, and is disposed to believe that the direct influence of changes of law on the figures may be traced. " It seems quite apparent," he says (p. 150), "that the lines of statistics are curved in accordance with laws enacted just prior to the curves," and he enumerates fourteen cases of synchronous changes of law and figures, in which the former " may account " for the latter. These are not the only changes of law, be it observed, for he alludes to others which 4 g THE DIVORCE PROBLEM. presumably were not attended by a change of figures. He gives no clue to their number; but the importance of these negative cases is quite as great as that of the positive, and their existence must be regarded in balancing the evidence. The figures change so constantly and greatly, that an occasional coincidence of new law and corresponding change of statistics would be almost inevitable. A fundamental oversight is made in this part of Mr. Wright's discussion. A change of the law is found to coincide with a change in the total of divorces for the state, and the two are forthwith connected, regardless of the fact that very few changes of law can have affected all causes of divorce equally. To establish a connection between the two as even probable, the change in the number of divorces must be shown to occur solely or mainly in the classes affected by the law. A statute making habitual drunkenness a ground of divorce may be at- tended by an increase in the total, but the two facts are not proved to stand in relation, until the increase is shown to be solely or largely in divorces granted for drunkness. A law shortening the requisite period of desertion cannot have in- creased the divorces for adultery, nor can a law making cruelty a cause have increased the divorces for desertion. Equipped with this simple touchstone, let us examine the fourteen cases given in the Report as instances and proofs of a change of law causing a change in the figures. They may be grouped as fol- lows : two introductions of new causes, South Carolina and Ala- bama; two repeals of causes, South Carolina and Connecti- cut; five modifications of causes or procedure, Dakota, West Virginia, Massachusetts, and two in Vermont ; five more com- plex changes, Maine, Indiana, Mississippi, Colorado and Utah. § 25. Changes introducing Causes. South Carolina passed a law in 1872 allowing its courts to grant divorces. From the fact that 98 per cent, of those granted involve either adultery or desertion, we infer that these were the grounds admitted. The figures show noth- THE DIVORCE PROBLEM. 4 o ing like a rush to her courts for relief. They increased slowly and irregularly for seven years, and at the end of the period had risen only to 39, while the annual average was 25. Alabama introduced habitual drunkenness as a ground of divorce in 1870. Another law of uncertain significance, passed the same year, was repealed in 1873 and Mr. Wright thinks the two " would perhaps account for the increase of divorce from 1870 to 1873." That increase was 29; the increase of divorces for drunkenness was 1. No divorces for this cause were reported in 1870, and 1 in 1873. If all the cases in any way involving drunkenness as contributing cause be included,, the argument is not materially strengthened. Such cases show an increase of 3, from 2 to 5. Nor are the two years exceptional. The figures for a decade only confirm the con- clusion. They are given below, the vertical line marking the change of law : •67 '68 '69 '70 Drunkenness alone • o o o o Drunkenness and other causes 4 3 3 2 The total increase for the decade was 105. What influence, did this law have in causing it? § 26. Changes repealing Causes. The effect of increasing the facilities for divorce may be gradual ; the effect of diminishing them must be immediate. The community may be slow in learning that the door has been opened more widely, but if it be swung shut a little, some are excluded from the start, or the effect is nil. There- fore a decrease of divorce beginning two or three years after the passage of a restrictive law cannot be set down to its agency. South Carolina illustrates the difference between the two kinds of legislation. The slow increase for seven years has been mentioned, but when divorce was prohibited in 1878 the number fell to nothing at once. Connecticut affords the only other example of a simple re- 4 '71 '72 '73 '74 '75 '76, O 2 I I I I jo THE DIVORCE PROBLEM. peal of a cause. The law in that state permitting divorce for " any such misconduct as permanently destroys the happiness of the petitioner and defeats the purposes of the marriage re- lation," was repealed in 1878. The repeal was approved in March, but the number of divorces for that year was the same as for the year before, 412. It might be said that the law checked the increase in Connecticut, and it is true that there had been an increase of 32 the year before, but this was ex- ceptional, and for two years before that the decrease had been marked. Here are the figures: '74 '75 '76 '77 '78 '79 '80 '81 '82 '83 '84 '85 '86 53! 498 380 412 412 325 346 428 401 423 344 398 420 Is any influence of the law here discernible? To be sure there is a considerable decrease, 2 1 per cent., a year later, but this is almost matched by the decrease of 19 per cent, in 1884, and considerably surpassed by that of 1876, 24 per cent., al- though in neither case was there a change of law. And the diminution comes too late to find a clear explanation in the legal change. The transfer of divorces for misconduct into other categories may be set forth as follows : 1877 Divorces involving misconduct.. 242 All others 170 1878 1879 1880 1881 1882 143 25 3 I o 269 300 343 427 401 412 412 325 346 428 401 The law immediately diminished the divorces for misconduct with or without other causes by 99, but there are just 99 more for other causes. It cannot be admitted that " the decrease is sufficient to show the influence" of the change of law. § 27. Modification of Causes or Procedure. Dakota in 1881 reduced the time necessary as basis for a complaint for desertion from two years to one, and the figures are here given : 88 1 1882 1883 1884 3° 74 H 98 47 49 68 74 THE DIVORCE PROBLEM. c r 1878 1879 1880 Divorces involving desertion . . 8 18 35 All others 9 12 37 The increase is clearly perceptible, but not very permanent. The conditions in that case changed so rapidly in these years that inferences are somewhat unsafe, but whatever weight is given to the instance should probably lie in favor of Mr. Wright's view. West Virginia passed a law in March, 1882, requiring resi- dence for a year previous to filing suit. Prior to that date resi- dence at the time had been sufficient. Divorces diminished from 204 to 176. The law was evidently aimed at divorces to non-residents, and the true measure of its influence must be found in the diminution of such divorces. The only datum is the number of divorces to parties married outside the state. Those married and divorced in West Virginia probably lived there in the interim. Now the divorces to parties married in that state diminished 13 per cent., while those to persons mar- ried elsewhere diminished 13^ per cent. Mr. Wright ingeni- ously suggests that the increase the year previous may be accounted for by the general knowledge that such a law was under discussion, so that cases were hurried through to take advantage of the laxer provision. This is sufficiently answered by noting that the divorces of parties married in West Virginia increased 104 per cent, that year, while those of parties married elsewhere increased 69 per cent. In this state, then, no influ- ence of legislation is traceable. Vermont made a like change in January, 1879. Divorce was refused to parties who had never lived together in that state, and restrictions were imposed upon divorce for causes arising outside. The decrease for which this legislation is as- sumed to account proves on analysis to be a decrease of 36 per cent, in divorces of Vermont marriages and a decrease of only 26 per cent, in those of marriages solemnized elsewhere. A second amendment in 1884 aimed to secure the presence 52 THE DIVORCE PROBLEM. of the defendant by compulsion if necessary. This was at- tended by a marked decrease in all the causes, but whether the relation was other than a coincidence cannot be statistically determined. Massachusetts in 1 873 reduced the time of desertion neces- sary to give ground for divorce from five years to three, and in 1875 raised it again. Mr. Wright says, "probably the large number in 1 874 was due to " this change. The total increased from 337 in 1872 to 61 1 in 1874, but, unfortunately for Mr. Wright's position, the increase under the head of desertion was only 69 per cent, while in those not involving this cause the increase was 92 per cent. Again the hypothesis fails. § 28. More Complex Changes. Maine amended its law in 1883 by forbidding the parties to marry within two years after the decree, and repealing the authority of a Justice of the Supreme Court to decree a divorce "when in the exercise of a sound discretion he deems it reasonable and proper." The change was attended by a sudden and marked decrease in all causes. A tendency to revive followed, but the rate did not reach its former propor- tions. Mississippi relaxed her law in three directions in 1871. Habitual drunkenness and desertion were made grounds, and the period of desertion required was diminished one year. The results are expressed in this table: Divorces for 1867 1868 1869 1870 1871 Desertion 22 30 26 27 32 Drunkenness o 1 1 o o Cruelty 1 o 2 1 2 Other grounds 26 28 46 57 71 Total 49 59. 75 85 105 Percentage of in- crease 20 27 13 24 1872 1873 1874 1875 1876 68 67 91 66 82 1 4 3 ' 4 12 15 9 20 15 79 83 73 84 71 170 169 176 171 172 62 o 4 — -3 o The increase for the single year was marked, more than twice as great a percentage as for any previous year, but it then stopped almost entirely. THE DIVORCE PROBLEM. , , Indiana introduced several amendments in 1873. One, a repeal of the clause, " for any other cause the court deems proper," resulted in an immediate, marked and permanent de- crease in the number of divorces reported as decreed " for cause unknown." When the judge was deprived of this discretionary power he became bound to refuse a petition, unless the case was brought under some statute. The other changes fixed at two years the duration of " deser- tion and neglect to provide" requisite for a divorce. The suits involving these causes diminished 24 per cent, others 17 per cent.; but the former began immediately to increase more rapidly than the latter, so that in 1875 they were 5 per cent, below the number in 1 872 ; those for all other causes were then 3'per cent, below. Colorado has a unique law upon desertion. The general provision of the frontier states allowing divorce for desertion continued one year is supplemented there by a law allowing it on this ground when husband or wife has left the state with- out intention of returning. Originally this law was in favor of the wife only, but in 1 88 1 its privileges were extended to a deserted husband. The result of such a change would natu- rally be to increase the divorces for desertion granted to hus- bands; but as a matter of fact those to wives increased more rapidly. So this change was of no perceptible effect. Two others made at the same time introduced the cause, neglect to provide, and changed the duration of habitual drunkenness necessary; but the increase under these two heads was only 18, while the total increase was 112. Therefore, the legal changes in Colorado had little influence on the increase. The last and most interesting case is Utah. That territory had a very lax law, but until 1875 not a large number of di- vorces. In that year divorce lawyers in New York, Cincinnati and Chicago began to avail themselves of the opportunity thus afforded, and to hurry eastern divorces through the Utah courts. Statements of an intention to take residence in Utah were c 4 THE DTVORCE PROBLEM. enough to secure jurisdiction, and affidavits of incompatibility of temper, to justify a decree. This continued for four years, when the Utah law was amended, bona fide residence for a year before the commencement of proceedings required, and the incompatibility of temper clause repealed. The apparent influence of this legislation is shown in the following figures giving the annual number of divorces for nine years: '72 '73 '74 1 ' 7S '76 '77 '78 | '79 '80 IOO 134 149 1 295 709 914 298 1 122 "5 It seems on the face of the figures that here was a reduction of about 800 divorces a year caused by the passage of the law. But was it really so? These divorces from eastern cities were sent to Utah, because that was the line of least resistance. When that channel was closed, the current may have returned to its original bed and flowed through the courts of those cities. If this hypothesis is correct, there should be during the period of numerous Utah divorces a decrease in the di- vorces of these eastern cities, followed by an increase when the law of that territory was amended. The following table has been constructed to test the hypothesis : Annual aver- Annual aver- Annual aver- Annual aver- age 1871- -74- age 1875-78. age 1879-82. age 1883-86. 468 342 603 722 New York City. 217 164 239 252 Utah (divorces of persons mar- ried at a dis- 31 439 31 34 Unfortunately the Cincinnati records were destroyed in 1884, but it can hardly be doubted that these figures comple- ment each other. This becomes still clearer when it is found that the rest of Illinois and the rest of New York maintained a steady increase through these sixteen years. The corre- sponding averages for the rest of Illinois are 1,248, 1,333, 1,568, 1,722, and for the rest of New York, 435,464,604,692. Possibly a rough measure of the number of divorces deflected from these cities may be obtained as follows. Assume the THE DIVORCE PROBLEM. cc real number of divorces to Chicago and New York parties for 1875-78 to have been a mean between the given numbers be- fore and after, i. e. 536 and 228. The sum of these numbers is 258 less than the sum of the numbers in the table. Ob- viously about 400 divorces annually went to Utah from the East; and here are approximately 250 accounted for in these two cities, leaving about 150 for Cincinnati and other places. The efficacy of this law in reducing the total of divorces for the whole country is made doubtful as soon as the. facts are scrutinized with care. The result of the examination of the fourteen test cases se- lected by Mr. Wright may be briefly recapitulated. Five, Alabama, Connecticut, West Virginia, Vermont (law of 1878), and Massachusetts, are to be unconditionally rejected; five others, Mississippi, Indiana, South Carolina (law of 1872), Colorado and Utah indicate a slight, temporary, or questiona- ble influence of legislation, two, Maine and Vermont (law of 1884), are not subject to closer determination from the figures, and in only two, Dakota and the repeal of the divorce law by South Carolina in 1878, is there clear evidence of a considerable influence on the figures. The former might be explained either as a mere coincidence or by reference to the very great increase of population in that state in that year, making a marked in- crease of divorce natural, and the case of Utah might be con- sidered as throwing light on that of South Carolina. It is indeed possible that some of the 25 divorces a year, shut off in that state in 1878, appeared in adjoining states, but neither of these explanations is satisfactory. Yet after giving due weight to these exceptions it must be admitted that the influence of law, if not nil, is at least much less than commonly supposed. § 29. European changes from Separation to Divorce. One objection may still be urged. Other states, it may be said, have not been in earnest; they have been content with make-shifts. South Carolina alone has courageously gone to the root of the matter, and her legislation has been a success. c6 THE DIVORCE PROBLEM. The true inference is, follow her example, strike deep and hard, and the blow will tell. Few are bold enough to advocate a re- peal of all divorce laws, but a large number urge a return to the policy of the Catholic church, refusing in all cases the right of remarriage. The postulate on which this argument rests, that people divorce in order to remarry, has already been refuted, (§ 1 3) but the effect of such legislation on divorce demands statement here. No American state has made the change; therefore, it is necessary to resort to European statistics. After Alsace-Lorraine was annexed to the German empire, the Ger- man divorce law was introduced in 1874. Prior to that time the French law had allowed only separation, without the privi- lege of remarrying. But this change of law did not result in a marked increase of suits. The figures for 1870-73 are lack- ing, but the following, taken mainly from Bertillon, indicate how little the regular increase from 1850 to 1886 was affected by the change in 1874. 74-75 4.52 1-1851-55 '56-'6o '6i-'65 '66-'6g '7o-'73 1 '• Separations or divorces J ' -> 2.48 2.89 3.83 wanting | to 1000 marriages . . J i-tio- •81 '82 '83 •84 '85 '86 I 6.80 10.3 1 1.3 12.6 12.1 13.3 11. 1 Divorce has increased with considerable rapidity in Alsace- Lorraine, but so has separation elsewhere; and that increase has been especially rapid since 1 87 1 in all parts of the Conti- nent. The change from separation to divorce, however, did not much accelerate the increase. The experience of France seems to prove the opposite. Divorce was introduced in 1 884, and the number rose from 3000 separations per annum to 6000 divorces. This is a marked instance of the influence of legislation on divorce. But it is not in conflict with the views of the party which holds law to be a minor factor. Indeed, the result was predicted by Bertillon in his book published to advocate the French law. He also pre- dicted that the divorce rate would gradually diminish until it was near the rate of separation before the law, and then THE DIVORCE PROBLEM. e- slowly increase. It is too soon to state whether this predic- tion also will be verified, but the reasoning on which it is based is. clear. A large number of couples in France as everywhere are living in estrangement and separation. Some have formed new and illegal unions. Some have obtained a decree of sepa- ration, but as this would not legalize a second marriage, others have deemed it a superfluity. Then a divorce law is passed, and thousands who have already obtained a separation have the decree converted into a divorce ; other thousands now see a reason for going to court where before none existed. After a few years this current will cease; all who have long desired a divorce will have obtained it, and only the new cases will be carried into court. Divorce was introduced into France by the Revolution, modified by the Code Napoleon, abrogated at the Restoration. The statistics for so early in the century as 1816 are not very trustworthy, but they indicate no marked diminution of con- jugal quarrels as a result of the transition from divorce to sep- aration. On the contrary, from 181 1 to 18 16 there were 72 divorces to 100,000 marriages; from 18 16 to 1 8 19 there were 75 separations to 100,000 marriages. The slight influence of legislation may be exhibited more clearly by comparing Belgium with France. The Code Na- poleon was introduced into the former country early in the century, and has remained in force ever since. Divorce and separation are both admitted, and divorce by mutual consent allowed. Until the recent amendment of her law, France allowed only separation, and that not by mutual consent. The two countries are quite similar in their religion and general conditions. Yet for the decade 1872-81 to every 100,000 couples Belgium had only 24 divorces and separations com- bined, while France had 33 separations.* Or, if a fairer com- parison be made between Belgium and the two departments of French Flanders just over the border and most similar to * Die Bewegung der Bevolkerung in der Schweiz im Jahre 1885, Beilage I. eg THE DIVORCE PROBLEM. Belgium in all respects, the result is that in each there were five legal decrees to iooo marriages, although in France the decrees were for separation, and in Belgium for either. That is, after living for three-quarters of a century under laws as totally different as the Catholic and Protestant theories, the two regions are found to have a perfectly identical rate. Can it longer be doubted that a return to the policy of separation would not solve our problem? § 30. Laws affecting the Expense of Divorce. The one efficient means of reducing the number of divorces by law is to make them expensive. The - evidence demanded will be furnished, but the money may not. The English law illustrates this. Prior to 1858 divorces were granted only by Parliament, the House of Lords sitting as a court. The ex- pense was enormous, probably some thousands of dollars, for two suits at law must be won by the plaintiff before a hearing was given in Parliament. The much-discussed divorce law of 1857 simply created a court to transact the business more rapidly and cheaply. The procedure was substantially the same, the causes for which divorces were granted were un- altered; and the reason that divorces have increased to some hundreds a year, when before only one or two were granted, must be found in the difference of expense. Yet even now, with only one court for England and Wales, the cost of carry- ing a suit through must be some hundred dollars at least, and in this may be found the fundamental reason for the small number of divorces in that country. France has pursued the opposite policy. A law was passed in 185 1 allowing those unable to pay the expense of a suit for separation, to plead without cost. It resulted in a marked in- crease in the number of applications. The obvious objections to having one system of law for the rich and another for the poor, make discussion of this method of restricting divorce unnecessary. THE DIVORCE PROBLEM. 50 §31. The Uniform Law in Switzerland. Uniform laws on divorce are needed in the United States, but so are uniform laws on bills and notes and various other subjects. The divorce problem would hardly be touched by uniformity. The present differences between the states would continue to exist almost unaffected.\/As they were not created by law, so they cannot be abolished by law. If the reader be not already persuaded, the example of Switzerland may con- vince him. That country is remarkably like our own in its divorce rate, but the differences between its cantons are greater than those between our states. Formerly, each canton con- trolled the subjects of marriage and divorce, but in 1874 the Constitution was revised and the Federal authorities em- powered to pass a national marriage and divorce law. Such a\ law went into effect January 1st, 1876, and a recent volume of Swiss statistics summarizes the results of ten years of the new law. The differences between the cantons have been very slightly affected thereby. The following table,* based on those results, may profitably be compared with table IX., § 20, the basis of the two being the same. TABLE XI. Divoree\Rale of Swiss Cantons under Uniform National Law for Decade 1876-1885. Zug 83 Orisons 1 1 5 Basel, country 159 Aargau 1 62 Solotburn 168 Vaud 180 Base), city 193 Neufchatel 198 Switzerland 208 Berne 223 Unterwalden o. d. W. . 9! Uri 10 Valais 15 Ticino 22 Unterwalden n. d. W.. 24 Schwyz 42 Lucerne 62 Freiburg 66 Appenzel 1 , Inner Rhodes 71 St. Gall 245 Thurgau 326 Geneva 336 Schaff hausen 336 Glarus 339 Zurich 379 A p p e n z e 1 1 , Outer Rhodes 440 After ten years of a uniform law, Appenzell, Outer Rhodes, has forty-nine times as much divorce as Unterwalden o. d. W., while with all the divergences of law in this country the differ- ences of rate are much less. * Compiled from Die Bewegung der Bevfilkerung in der Schweiz im Jahre 1885. Beilage I. 6 THE DIVORCE PROBLEM. §32. Legal Restrictions upon Marriage. Certain students of the divorce problem have advocated re- strictions* by law upon early or improvident marriage. For example, Prof. Robinson, of Yale University Law School, says:* "No person should be marriageable under the age of 21, and a marriage ceremony celebrated between persons either of whom is under age should be ipso facto void." Four thousand eight hundred and fifty-five married persons under the age of twenty were living in Massachusetts in 1885. To these must be added a large and indeterminate number of married persons between twenty and twenty-one, in order to ascertain the number of marriages which would be declared void by such a law as is proposed. How large a proportion of these nearly ten thousand people (for in the great majority of instances only one of the two is under twenty-one), would have remained virtuous and continent in the face of a law for- bidding marriage? Each reader must judge for himself, but the experience of Bavaria may aid him in forming a conclusion. In that country the local authorities were empowered to refuse marriage to such as could not give reasonable evidnece of ability to support a family, in short to paupers. The num- ber of marriages decreased rapidly, but parallel with this de- crease went a large increase in the number of illegitimate births, until they reached a total of nearly one-fourth the births in the kingdom. Disturbed by this result, the legisla- ture changed the law in 1861, and at last entirely repealed it. The annual number of marriages leaped at once from 38,000 to 59.000, and remained abnormally high for several years, thus proving the number of persons who were glad to marry when the law allowed. Simultaneously with this came a marked decrease in the number of illegitimate births. Similar results have been obtained elsewhere. Therefore legal re- strictions upon marriage cannot be deemed a satisfactory method of checking divorce. * The Diagnostics of Divorce, Journal of Social Science (Am. Ass.), xiv. (1881), p. 136. THE DIVORCE PROBLEM. 6 1 § 33. Summary of Results. The proposed modes of reducing divorce by law may be grouped as restrictions on marriage, restrictions on divorce, restrictions on remarriage. Restrictions on marriage reduce the number of marriages, and thus ultimately the number of divorces. By excluding the poorest or lowest classes they may do this, not only in comparison with the population, but even in comparison with the married couples. The attendant evils are so great never- theless, as to make such restrictions unwise. Restrictions on remarriage would probably not reduce the number of divorces. The statistical evidence obtainable indi- cates that divorces are not sought in order to remarry. All the objections to restrictions on marriage weigh with equal force against restrictions of this class. Furthermore, countries which have similar conditions but different laws on remarriage, present rates of divorce or separation practically identical. Restrictions on divorce exert a minor influence on the rate. The three belts are not explained by differences of law ; simi- larity of conditions and sentiment rides roughshod over diver- sities of statute in our own adjacent states, while in Switzer- land diversities of condition hide any effects of a uniform law. The single efficient means of reducing divorce by law, neglect- ing as unadvisable and impracticable the South Carolina method, is to make it expensive. This is open to all the ob- jections against restrictions on marriage. It makes one law for the rich and another for the poor. The- conclusion of the whole matter is that law can do little. Agitation for a change of law may educate public opinion. It may even be the most efficient and powerful means of edu- cation. Such effects no statistics can measure, and therefore in a paper like this the educative influences of law must be neglected, but the immediate, direct and measurable influ- ence of legislation is subsidiary, unimportant, almost impercep- tible. PART III. CAUSES AND REMEDY. § 34. Basis and Character of Conclusions. In so complex a social plenomenon as this, the line of statistical demonstration is short, and its limit soon reached. While following it, however, opinions may gradually develop, which statistics can neither justify nor gainsay, and a few, struck out in the course of the present study, will be appended here. Some are hardly more than hypotheses to be verified, modified or retracted on further study; others, perhaps, may rank as probabilities; but all alike are offered merely as sug- gestions. § 35. Two Conceptions of Marriage Law. A fundamental antithesis underlies the marriage laws of the Christian world. How far the antithesis is really one of race between Teuton and Celt, it would be hard to decide. Ordi- narily it is treated as one of religion between Catholic and Pro- testant. The Catholic theory is that marriage law should recognize, embody and hold fast to, a moral or religious ideal, unmodified by consideration of the needs and moral condition of the com-, munity to which the law applies. This may be termed the idealistic theory. The realistic theory of Protestantism lays more stress on the moral standards of the community, and finds greater difficulty in solving the divorce problem with a simple, " Thus saith the 'Lord." The different conceptions of marriage law may be illustrated by the different treatment of celibacy in southern' and northern (62) THE DIVORCE PROBLEM. g, Europe. Celibacy may be a moral duty for certain natures. In that belief the Catholic .Church has attempted to enforce the fulfillment of vows of continence with legal pains and pen- alties. Protestantism admits that celibacy may be sometimes a duty, but insists that the attempt to enforce it by law is not beneficial or wise. Only at times of special enthusiasm can the sexual instinct be permanently repressed. For such times the law is not necessary, and during the intervals it is injur- ious. Similarly, the moral or religious ideal of marriage is a life- long union between man and wife. Yet, as a matter of fact, that union does terminate and all relations between the parties cease. Whether the law shall recognize this de facto termina- tion of marriage depends upon the conception of the function of law held by the community. If that function be to hold up an ideal, no falling short of the ideal can be admitted. Now the point here made is that since the Reformation the realistic view of marriage law has been constantly supplant- ing or modifying the idealistic. Whether this change has been attended by the lowering of the ideal of marriage, would be hard to decide; but it certainly has exerted much influence on the spread of divorce, or the legal recognition of the termi- nation of a marriage. § 36. The Popularization of Law. During the Middle Ages l aw was a personal privilege. For gentuneT Ttegal TbrmF oT procedure continueci~so intricate and expen sive that the b enefits of laTvaccrueTonTyToth^wis^oi^ wealtny. Along ~with tne e xten^ionoT"the*"s^jrrrage in modern time's has cova^^j^mQ^^s^^^^^^^^^^^V^^^ 5 - Thole classes have been admitted to courtrliaT^e1 : eioT : merly excluded by the efficient practical prohibitions of ignorance and poverty. The change in the position of the negro, effected by his emancipation, is but a single striking illustration of what has been going on constantly as a result, on the one hand, of laws simplifying procedure and diminishing the ex- 6 4 THE DIVORCE PROBLEM. pense of litigation, and, on the other, of the better education of the community in matters of law. This education is con- ducted largely by the newspaper press of the country. Many a man would live in ignorance that such a thing as divorce ex- isted, were it not for the conspicuous mention of trials in his morning paper. Thus the law has become a weapon of offense or defense for a very much larger part of the population than could use it even so recently as fifty years ago. In consider- ing the rate of increase estimated from the figures, (§§ 3—1 1,) this must be borne carefully in mind. Imagine society as a huge pyramid in which the position of each individual is determined by his knowledge and wealth. Imagine a horizontal plane intersecting the pyramid to repre- sent the divorce law of the community, and all persons above the plane as possessing so much knowledge and money that divorce is to them a theoretical possibility, while to those below it is not. If the plane be motionless, the rate of increase of divorce may be found ; but if it be gradually sinking towards the base of the pyramid, and making divorce a practical possi- bility to an increasing proportion of the whole number, this change must affect the calculation. Su ch a descent of the divorce plane has been in progress in this\country, apparently, tor the past twenty jvears. While it doesvnot invalidate the previous conclusions, it does influence them, perhaps materi- ally, and certainly renders untrustworthy any estimate for the future. § 37. Laxity in Changing and Administering the Law. There are three parties to every divorce suit, the husband, the wife and the public. The consent of the public to a de- cision usually desired by both the other parties is expressed by the judge's decree, given in conformity to the law which itself expresses, with more or less accuracy, the permanent opinion of the community. If the state be small and the population scanty, its consent is easily gained. Push the subdivision of a state to its ultimate atoms, the families, and we revert to the patriarchal period when state and family THE DIVORCE PROBLEM. 5, were identical. Then the will of the public and that of the head of the family coincided, and divorce required no sanction from a court. As society grew compact and better organized, the restraint of public opinion outside the family made itself felt, at first in the mere requirement of certain formalities, like a " bill of divorcement " or a prescribed formula. In new and sparsely settled communities, like those of our frontier states, there is an appreciable return towards the patriarchal con- dition. Public opinion is hardly formed, its restraint hardly felt; divorce becomes almost a personal and private matter. Judges, unrestrained by public sentiment, are lax in their de- cisions and legislatures in their enactments. The differences between the law of our Atlantic coast states and that of "Eng- land, and between the law of our eastern and western belts of states, illustrate the tendency to a relaxation of law manifested in all new communities. The divorce law of Canada and Australia would probably have been changed ere this, but for the restraining influence and, in the last resort, the veto power of England. If our common law had been under the exclusive protection of the National Congress, and not subject to modification by- the states, its provisions on divorce would have changed muchi less; if the common law of England might have been modi- fied by each county, the law of that country would, doubtless,, be much laxer. Denmark and Switzerland have the highest divorce rate in Europe, and Switzerland, till 1876, had sepa- rate laws for each canton. While the direct effect of a change of law is slight and ephemeral, its importance as a register of the public sentiment of the community is very great. The public sentiment of new and small communities changes much more rapidly than that of old and large ones, and, in the last resort, that sentiment determines the number of divorces. § 38. Age of Marriage. No direct connection between the age of marriage and the 5 66 THE DIVORCE PROBLEM. liability to divorce can be made out from the statistics. Yet it seems to be the rule that the communities in which early marriages are most common, are most free from divorce. Thus Prince Krapotkine tells us that in Russia "the peasants for the most part marry their sons at eighteen, and their daughters at sixteen," and the Russian peasantry are perhaps, with the exception of the Irish, the freest from divorce in Europe. Mr. Lecky, however, lays stress upon "the nearly universal custom of early marriages among the Irish peas- antry," as explaining the remarkable chastity of that people. § 39. The Emancipation of Women. The emancipation of women means the attainment of such legal recognition and support as enables them to use the law for their defence with as much readiness and freedom as do men. It involves an economical and mental independence of men, whether as fathers or as husbands. If the organization of society greatly hinders women from becoming self-support- ing, the wife endures many wrongs to which the ear of the judge would be open, simply because life apart from her hus- band is starvation. As women become able to earn a living income this economic bond is relaxed. The effect on con- jugal life is seen in the greater number of divorces granted to the wife in the northern states. (Table VIII.,§ 19.) Perhaps the New England woman is somewhat more emancipated, i. e., more independant mentally and legally, than the woman of the middle Atlantic states. A divorced woman may gain a livelihood either from her own labor or by a second marriage. Therefore it is natural to find divorce most frequent where a woman finds it most easy to earn her bread, and also where wives are most in demand. The latter is probably the case in the trans-Mississippi states, and may serve, in part, to explain the excess of divorces in the far west (Table IX.), the very large proportion granted to wives (Table VIII.), and the fact that in those states, suffering wives show less patience than husbands, § 19. THE DIVORCE PROBLEM. fa The emancipation of women is closely related to the average age of marrying. Women marry early in those com- munities where no other vocation is open to them than that of wife and mother. Only sixteen to twenty years of age when she passes out of the control of a father and mother into that of a husband, with no taste of freedom intervening, with a mind and character so unformed as easily to be brought into harmony with or submission to her husband's, with no way of escape open to her after marriage, whatever the law may say, what wonder that the peasant woman of Russia, Ireland or elsewhere shows little inclination to divorce! Where no actual choice is offered, divorce cannot be chosen. Between the cessation of strict parental discipline and the beginning of married life, a period of some length must intervene, in order that a woman may waken to her own equality and independ- ence. Thereafter, any marriage she may contract will be based on equality, not on subjection. The economic emancipation of women, in the forms it has thus far assumed, is attended by an assimilation of the work of wage-earning women to that of men. Marriage, however, is fundamentally grounded on the differences, physical, intellec- tual and moral, between the sexes. Consequently a marriage almost invariably recognizes and emphasizes these differences through varieties of work and function. So far as the train- ing of the two sexes prior to marriage has been identical, one or the other must be ill fitted for that life; so far as woman's work has become masculine, her ability to make and keep a home happy is diminished. This result appears most clearly in the wage-earning population, where a girl's marriage is less often a change from one home to another and more often a change from a factory into a home. It is in just these classes that divorce is most frequent. Some confirmation of this conclusion may be drawn from a report of the proceed- ings of the London Divorce Court quoted by Prof. Robinson.* ♦Diagnostics of Divorce, Journal of Social Science (Am. Ass.), XIV: (1881), p. 136. 68 THE DIVORCE PROBLEM. " The experience of the English divorce court tends to confirm the opinion that a great deal of the misery to which the working classes are subjected in their homes arises from the inability of women, when they get married, to render their homes comfortable and attractive. In other words, a great many of the women who get married are unfit for married life." Divorces are most frequent where women are most emanci- pated, and the percentage granted to the wife in such communi- ties is excessive. For the whole country the percentage to women is increasing. In the first year, 62 per. cent., in the last, 66 per. cent., were granted to the wife. § 40. Growth of Cities. The proportion of our population living in cities of over 8000 inhabitants has increased from about 16 per cent, in i860 to about 25 per cent, in 1890. The native urban population so the census of Massachussetts indicates, are more prone to divorce than the same class of population in the country. Hence the massing of that population in cities at the east and north has probably influenced the divorce rate, although the effect is masked by the presence of the foreign born urban population. § 41. Increase of Industrialism. The life of the race on this planet has been divided into four stages, each distinguished from the others by a predominant manner of obtaining food. They are the hunting and fishing stage, the pastoral stage, the agricultural stage, and the indus- trial stage. It has been further argued that permanent mono- gamous marriage makes its appearance with the dawn of the agricultural stage. However that may be, it is obvious that we are in the midst of the transition from the agricultural to the industrial stage. Not that hunting, fkhing, herding or farm- ing has passed away or is likely so to do. But the character- istic feature of the present day lies in none of these ; they have existed in the past; trade, commerce and manufacturing in anything like their present importance, have not. This advance into the industrial period is apparently attended by a modifica- THE DIVORCE PROBLEM. 69 tion in the relations of the sexes, and perhaps in the nature of marriage. Industrialism tends to eliminate all sexual differ- ences except the physical ones, and to reduce even these to a minimum. It centers in cities, has grown faster and farther at the north than at the south, and perhaps faster in New Eng- land than in the middle Atlantic states. Industrialism involves a close relation between all parts of a community. The newspaper and periodical press mediate the exchange of ideas and information, as the railway and tele- graph do of material products; the knowledge that divorces are obtainable is disseminated by the former, the opportunity for cheap and easy transportation to a more alluring field of work is offered by the latter. The number of divorces granted for desertion alon^jsjwo^ffihj^ofjhe whole, and nearly one-half involve this ground as principal or accessory cause. The pro- portionT'too, of divorces for this cause has steadily risen from 34 per cent, in 1867 to 40 per cent", in 1886. The mental and economic emancipation of women has weakened the ties bind- ing a wife to her husband ; the facilities for transportation and the reports of golden harvests in California, the Black Hills or Oklahoma have increased the motives for a husband to abandon home and wife. Industrialism involves a wider knowledge, action under more complex and changeable conditions, and thus a vastly greater mental activity. As the mental life increase's, the general dif- ferences of sex and the specific differences of persons are ac- centuated, and the dangers of quarreling increase. A savage finds one squaw about as serviceable as another in attend- ing to the few tasks of the wigwam; in our higher civili- zation dissensions ending in divorce may begin (Report, pp. 172-178), over a curly dog, or _a woman's ban gs, or a man's buttons, as well as over spiritualism or religion. This demand for higher mental life many fail to meet; they succumb under the strain. The growth of industrialism is at- tended by an increase of insanity, as well as of divorce. •jq THE DIVORCE PROBLEM. § 42. The Spread of Discontent. So closely related to the growth of industrialism as hardly to deserve treatment as a separate cause, is the spread of a spirit of restless dissatisfaction. Mr. Bryce has called this "the ^a ge of dis content," and this characteristic of the time m this country and Europe manifests itself in a theoretical question- ing and criticism of marriage, and, perhaps, in the weakening of its hold upon the community. Even more characteristic of the industrial age than the in- crease of insanity is the increase of suicide. This is the high- est and deepest expression of discontent with the sum total of conditions constituting one's life. Every suicide involves, for the individual concerned, and practically, not theoretically, a negative answer to the question, Is life worth living ? Like- wise every divorce involves, for one or both of the parties, and practically, an affirmative answer to the question, Is marriage a failure ? Therefore it is not surprising to find a close and constant relation between the statistics of divorce and of sui- cide. Both are much more common among Protestants than Catholics, in cities than in the country, among the Teutons than the Celts ; and both are rapidly increasing. The rates of divorce and of suicide in the particular countries show a close and constant relation, and the proportion of suicides among divorced persons is abnormally large. The discontent with marriage in its present form shown by the increase of divorce may be compared, not only with the discontent with life shown by suicide, but also, perhaps, with the discontent with the present organization of industry shown in strikes and combinations, and in visions of an entirely new economic organization or of an ideal socialistic state. The discontent in all its phases indicates a desire and hope of great change. For the suicide, even, must regard annihila- tion as a change and an improvement. Whether the changes will be realized or can be without a fundamental change in human nature, we need not attempt to decide; but ob- THE DIVORCE PROBLEM. y t viously the first condition of their realization is that they be desired. § 43. Two Ideals of the Family. Once before in the history of the world women have been emancipated legally and economically. It was in the days of the Roman empire. Wives were given by law and by custom' greater freedom than ever before; but they gained it by the sacrifice of family life, and under a theory that the wife was a member, not of her husband's, but of her father's household. The legal emancipation of women was attended by a loosening of the ties between husband and wife, a disregard of marriage vows, the moral degradation of the sex and of the community. The family was sacrificed to the wife. When Christianity obtained the power to legislate, the family was reinstated in supremacy^ and the wife again made legally subordinate to her husband. Since the Reformation the legal and economic emancipation of women has made fre- quent advances ; and the problem of the present day in this matter is : How shall the legal and economic emancipation of women be made compatible with the true interests of the family? A clue to the answer may, perhaps, be found in recognizing that there are two ideals of family life, and that, while the emancipation of women is impossible without de- stroying the one, as it did at Rome, it may be reconciled with the other. The Roman theory of the family is based on the complete supremacy of the husband and father, manus and f atria po- testas. This despotic theory, abandoned by the later Roman jurisprudence, was substantially restored by the canon law, and extended over Europe by the power of the ecclesiastical courts. The view that the wife was owned by her husband, and the not very dissimilar view that her legal personality during marriage was merged and lost in his, were natural and inevitable in times when property in persons, as slaves, was universally recognized and justified, and the original depen- ■j2 THE DIVORCE PROBLEM. dence of legal rights upon force hardly lost from sight. Slavery has been perpetuated into our own age. Its central idea, that property in persons is natural and legitimate, still lives and nourishes. Many a man feels himself the owner of wife and child, and treats them as his property. The kernel of the other ideal of marriage, the democratic ideal, is found in the Teutonic emphasis upon freedom and in- dividuality, and the Teutonic honor of womanhood. This ideal bases the family, not upon the despotic authority of a single head, but upon the consenting and harmonious wills of two equals; not upon manus, but upon marriage. The devel- opment and establishment of this ideal of the family have been retarded by the influence of the ecclesiastical courts denying the equality of women in marriage, and also by the prevalence of slavery and constant appeals to war and force as final arbiter of all disputes. Even down to the present time, this ideal has not made itself at home in the minds and hearts of all our people, and many a dispute between husband and wife is at bottom a clashing of the old and new, the despotic and demo- cratic theories. As a democracy requires more knowledge and political virtue than a despotism, so the successful and harmonious management of a family on a democratic basis of equality and delegated powers demands more fidelity and more adaptation than when a single will holds sway. § 44. The Remedy. The whole argument of this monograph has gone to show that legal provisions of whatever sort have little direct and permanent influence upon divorce. Restrictions on marriage, restrictions on divorce and restrictions on remarriage after divorce, have been tried in various places and at various times, and have proved of little effect. Law emphasizes rights; at all times of great change in the state or in law, that emphasis becomes excessive. The religious wars to assert the right to freedom of conscience, the French Revolution to proclaim the rights of man, the THE DIVORCE PROBLEM. 73 American Revolution to maintain the taxpayers' right to rep- resentation, and the Civil War to establish the slaves' right to be free, are familiar illustrations. The same is true of that peaceful revolution of the past hundred years by which women's rights have been increased. In all cases where too much has been expected of such legal reforms, disappointment or despair has resulted. The discontent Mr. Bryce finds charac- teristic of the age is, perhaps, in part explicable as the discour- agement resulting from the failure to regenerate the world on either side of the Atlantic by any sort of legislation. New rights can mean only new responsibilities and new duties ; un- less the one be accepted and the other performed, little or nothing is really gained. The whole ideal and tendency of our modern civilization are to teach every individual self-direction and self-government. No legal reform can do such work. The state, indeed, may do much, but not directly, through its laws. On the contrary, its main work must be as an educator of public opinion, and, like a wise teacher, it must set lessons not too difficult for the average capacity of the class. It is to hold up a standard of morality as far in advance of the average standard in the com- munity as possible, without bringing the law into contempt or disregard. So far as law contributes to this work of educa- tion and moral improvement, it may be of immeasurable value. Even greater are the opportunities afforded to the state by its schools and institutions of learning. Here it is brought into direct and intimate contact with the minds of citizens while young and plastic. Moral education on these subjects in our schools is sadly lacking. Obviously, it should relate, not directly to divorce, but to all the relations and duties of home life so constantly and sadly misunderstood. The ne- glect of these is the seed, and divorce only the fruit. The church or ethical society has, perhaps, greater power and better opportunities than the state for educating and puri- fying public opinion. The opportunities have long been sorely 74 THE DIVORCE PROBLEM. neglected, and public opinion allowed to degenerate. Our situation resembles in some degree that of the Roman Empire. Women, there, were emancipated by law and custom; again they have gained their freedom, legal and economic. A moral or religious reform, like that which came in Christianity, is needed to teach and enforce the new duties that have come with the new rights and new powers: new duties of wives to remain faithful to their husbands, though not compelled as formerly by law or economic dependence; new duties of hus- bands to treat their wives not as subordinates, but as equals. If it be admitted that the family is in a state of reconstruc- tion, that its old form is proving insufficient for meeting the new conditions, then a careful and thorough study of the sub- ject, in all its relations, is a prime necessity. Otherwise the education just urged will be incorrect and misleading. s Education in all the relations and admitted duties of home life, moral and religious reform to supply motives for a per- formance of the duties thus made clear, and study to deter- mine the changes in the family introduced by our new con- ditions and the attendant change in duties; such are the remedies that appear of permanent value. THE JOHNS HOPKINS PRESS, BALTIMORE, HARYUND. gMDIES IN pigT0^Y ^ND PffldTOJjS. Edited by HERBERT B. ADAMS. The Ninth Annual Series is now in progress. Subscription, $3.00 per year. ANNUAL SERIES. 1883-1890. Eight Series of the University Studies are now complete, and will be sold, bound in cloth, as follows : SERIES I.— LOCAL INSTITUTIONS. 479 pages. $4.00. SERIES II.— INSTITUTIONS AND ECONOMICS. 629 pages. 24.00. SERIES III.— MARYLAND, VIRGINIA, AND WASHINGTON. 595 pages. ^-CO- SERIES IV.— MUNICIPAL GOVERNMENT AND LAND TEN- URE. 600 pages. $3.50. SERIES V.— MUNICIPAL GOVERNMENT, HISTORY AND POLITICS. 559 pages. 13.50. SERIES VI.— THE HISTORY OF CO-OPERATION IN THE UNITED STATES. 540 pages. $3.50. SERIES VII.— SOCIAL SCIENCE, MUNICIPAL AND FED- ERAL GOVERNMENT. 628 pages. $3.50. SERIES VIII.— HISTORY, POLITICS, AND EDUCATION. 625 pageS. $3.50. EXTRA VOLUMES. I. THE REPUBLIC OF NEW HAVEN. By Charles H. Lever- more. $2.00. II. PHILADELPHIA, 1681-1687. By Edward P. Allinson, and Boies Penrose. $3.00. III. BALTIMORE AND THE NINETEENTH OF APRIL, 1861. By George William Brown. JSi.oo. IV-V. LOCAL CONSTITUTIONAL HISTORY OF THE UNI- TED STATES. By George E. Howard. Vol. I. — Development of the Township, Hundred, and Shire. $3.00. Vol. II. — Development of the City and the Local Magistracies (in press). VI. THE NEGRO IN MARYLAND. By Jeffrey R. Brackeit. $2.00. VII. THE SUPREME COURT OF THE UNITED STATES. By W. W. WlLLOUGHBY. Jl.25. VIII. THE INTERCOURSE BETWEEN THE UNITED STATES AND JAPAN. By Inazo Nitobe. $1.25. IX. STATE AND FEDERAL GOVERNMENT IN SWITZER- LAND. By John M. Vincent. (In Press.) The set of eight volumes is now offered in a handsome library edition for $24.00. The eight volumes, with seven extra volumes, "New Haven," "Baltimore,'' "Philadelphia," "Local Constitutional History," Vol. I., "Negro in Maryland," "U.S. Supreme Court," and "U.S. and Japan," altogether fifteen volumes, for $33.00. The seven extra volumes (now ready J will be furnished together for $10.50. A Quarterly Review for the Consideration of Social and Economic Questions. IN CONNECTION WITH THE OXFORD UNIVERSITY BRANCH OF THE CHRISTIAN SOCIAL UNION. The Economic Review will be concerned chiefly with the moral and social bearings of economic problems ; but it will also include more technical articles dealing with special aspects of our industrial system, or treating of the historical condition and development of some particular period, and it is hoped that no small measure of support may be given to it by the increasing number of people who are feeling the burden of responsibility in regard to the stress of existing social problems. The Economic Review will be sent post-free for Ten Shillings a year, paid in advance; the price per copy being Three Shillings. 1. THE QUESTION OF POPULATION. The Rev. and Hon. Arthur Lyttleton, M. A. 2. RODBERTUS-JAGETZOW AND SCIENTIFIC SOCIALISM. Prof. Emile de Laveleye. 3. SOCIAL CONDITIONS IN A NEW ENGLAND. The Right Rev. Alfred Barry, D. D. 4. THE JOINING OF ISSUES. T. Mackay. 5. THE AMERICAN COPYRIGHT BILL. C. J. Longman, M. A. 6. FREDERICK DENISON MAURICE A CHRISTIAN SOCIALIST. His Honour Judge Hughes, Q. C. 7. GROSS' GILD MERCHANT. The Rev. W. Cunningham, D. D. REPORT ON SOCIAL LEGISLATION IN THE UNITED STATES FOR 1889 AND 1890. Prof. Richard T. Ely, Ph.D., and L. S. Merriam, B. S. NOTES AND MEMORANDA. Legislation, Parliamentary Inquiries, and Official Returns. Edwin Cannon, M. A. REVIEWS. PERCIVAL &» CO., 34 King St., Covent Garden, London. NEW YORK: BERU!t: JAMES POTT & CO. POTTKAMMER & MUHLBRECHT. Subscription, $2.50. Single copies, 75 cents. Subscription, 12 marks. Single copies, 3 marks . Systematic Political Science BY THE UNIVERSITY FACULTY OF POLITICAL SCIENCE OF COLUMBIA COLLEGE. The University Faculty of Political Science of Columbia College have in preparation and intend to publish a series of systematic works covering the entire field of political science proper and of the allied sciences of public law and economics. The method of treat- ment will be historical, comparative and statistical ; and it will be the aim of the writers to present the latest results of institutional develop- ment and of scientific thought in Europe and America. Each work will be indexed by subjects and authors, and the last volume will contain a topical index to the entire series. The series will consist of the following nine works : Comparative Constitutional Law and Politics. By John W. Burgess. Comparative Constitutional Law of the American Com- monwealths. By Frederick W. Whitridge. Historical and Practical Political Economy. By Richmond M. Smith. Historical and Comparative Science of Finance. By Edwin R. A. Seligman. Comparative Administrative Law and Science. By Frank J. Goodnow. International Law. By Theodore W. Dwight. Historical and Comparative Jurisprudence. By Munroe Smith. History of Political Theories. By William A. Dunning. Literature of Political Science. By George H. Baker. The first of these works was published early in 1891, by Ginn and Co. The entire series will probably be completed within three years.