War Damage in the Nanking area Dec. 1937 to Mar. 1938

Cover

WAR DAMAGE IN THE NANKING AREA

December 1937 to March 1938

URBAN AND RURAL SURVEYS

Cover 2

WAR DAMAGE IN THE NANKING AREA

December 1937 to March 1938

URBAN AND RURAL SURVEYS

Foreword

Foreword

IN the course of its efforts to aid the impoverished and troubled people of Nanking and adjacent districts, the Nanking International Relief Committee early felt the necessity of discovering with reasonable accuracy their true economic position.

What were both the extent and the nature of their losses?

How far impaired are their opportunities and abilities for making a living? What supplies of food may be expected from the farms of this area?

What are the fundamental deficiencies or obstacles that stand in the way of normal economic effort on the farms and in the city?

Such questions were basic to any sound consideration of policy and methods in relief.

The only good way toanswer them was to go out to seek the facts.

The Nanking International Relief Committee here makes known the results of its inquiries primarily for the information of those concerned with the practice and the support of relief work in this and other areas; secondarily, for the wider public which is or should be concerned with the ravages of warfare among civilians, in whatever country.

Our own position is humanitarian, without regard to the nationality of war victims.

In this report we seldom use the terms "Chinese" and "Japanese", and consider persons simply as farmers,housewives or children.

The International Committee is aware, however, that statements have been published by Chinese, putting upon the Japanese an exclusive and exaggerated blame for the injuries to the people of the Nanking area; likewise that statements have been published by Japanese, charging the Chinese with burning and looting which they themselves benevolently checked.

In order to guard against controversial misuse of the present report, we feel it necessary to make a brief factual statement as to the causation of the injuries listed.

The burning in the municipal areas immediately adjoining the walled city of Nanking, and in some of the towns and villages along the southeasterly approaches to Nanking, was done by the Chinese armies as a military measure - whether proper or improper, is not for us to determine.

A very small amount of damage to civilian life and property was done by military operations along the roads from the south-east, and in the four days of moderately severe attack upon the city.

Practically all of the burning within the city walls, and a good deal of that in rural areas, was done gradually by the Japanese forces (in Nanking, from December 19, one week after entry, to the beginning

I

Foreword2

of February). For the period covered in the surveys, most of the looting in the entire area, and practically all of the violence against civilians, was also done by the Japanese forces - whether justifiably or unjustifiably in terms of policy, is not for us to decide.

Beginning early in January, there gradually developed looting and robbery by Chinese civilians; and later, particularly after March, the struggle for fuel brought serious structural damage to unoccupied buildings.

Also, there has latterly grown up in the rural areas a serious banditry which currently rivals and sometimes surpasses the robbery and violence by Japanese soldiers.

In some portions of our report, these elements of causation can be distinguished.

From a humanitarian point of view, we venture merely to point out that losses to life and property from actual warfare are shown by these surveys to be one or two per cent of the total.

The rest could have been prevented if both sides had wished to give sufficient consideration to the welfare of civilians, including reasonable protection by military and civilian police.

The International Committee which authorized these surveys had within its membership a trained sociologist, Dr. Lewis S. C. Smythe Professor of Sociology at the University of Nanking, who not only had general experience in survey methods, but also had taken a responsible part in two earlier surveys of calamities in this region.

These inquiries were: the Economic Survey made on behalf of the National Flood Relief Commission by the Department of Agricultural Economics of the College of Agriculture and Forestry of the University of Nanking (report published by Professor J. Lossing Buck, Director, under the title "The 1931 Flood in China"); and the Survey of the Rural Areas Affected by the Shanghai Hostilities (1932), made at the request of the Minister of Finance by the same Department of Agricultural Economics (unpublished).

Both these surveys were for the purpose of ascertaining actual needs as against vague or tendentious reports from local officials.

The accomplishment of the present surveys is largely dependent upon the unusual abilities and energies of Dr. Smythe, even though he has not been able to give full time to them while acting both as Treasurer and as Secretary of the International Committee.

Full acknowledgment is due to the surveys mentioned above, which were drawn upon both for points of method and for check or comparison of results.

Likewise to the vast survey recently completed under Professor Buck's direction, and reported in his book, Land Utilization in China, with supplementary Atlas and volume of Statistics.

M. S. BATES

II

Contents

Contents

Page

FOREWORD | i

List of Tables

List of Tables

Table

1. Families Studied and Estimated Population by Section of City

2. Age and Sex Distribution of Population Studied Shown in Percentages

3. Family Composition by Section of the City, Shown in Percentages

4. Number and Cause of Deaths and Injuries, by Date

5. Sex and Age of Deaths, Injuries by Military Violence, and of Persons Taken Away, Shown in Percentages

6. Former Employment and Earnings, by Occupation

7. Present Employment and Earnings

8. Sources of Cereals During Preceding Week, by Section of the Shown in Percentages

9. Losses of Buildings and Movable Property for Families Remaining in Nanking, by Cause

10. Losses of Buildings and Movable Property for Families Remaining in Nanking (by Original Address) According to Type of Building, Cause and Section of the City, Shown in Percentages

11. Number of Buildings Damaged or Looted, According to Type of Building Cause and Section of the City, Shown in Percentages

12. Losses of Buildings and Contents, According to Section of the City

13. Losses of Buildings and Contents, by Item and Cause

14. Losses by Damage and Looting of Buildings and Contents According to Type of Building, Cause and Section of the City, Shown in Percentages

15. Number of Buildings Damaged or Looted on Main Business Streets,According to Type of Building, and Cause, Shown in Percentages

16. Losses by Damage and Looting of Buildings and Contents on Main Business Streets, According to Cause, Shown in Percentages

17. Population and Cultivated Area of Five Hsien of Ningshu

18. Value of Farm Losses

19. Average Losses per Family

20. Losses of Farm Families with Detail of Items

21. Estimate of Volume of Winter Crops

22. Total Seed Requirements for Spring Planting

23. Migration and Labor Supply

24. Sex and Age of Deaths

25. Number and Causes of Deaths (During 100 Days Covered by Study)

26. Damage to Buildings

27. Losses of Labor Animals

28. Losses of Implements

29. Losses of Stored Grain

30. Area Planted to Winter Crops

31. Percent of Area Planted to Winter Crops Totally Destroyed

32. Expected Per Cent of Normal Winter Crops in Planted Area not Totally Destroyed

Page 1 INTRODUCION

INTRODUCTION

ORGANIZATION AND METHOD*

The International Committee's Surveys were really two, though each of them was compound.

The City Survey was essentially an inquiry among families resident in Nanking, supplemented by an investigation of all buildings unoccupied as well as occupied, and also separating for special attention as food-producers the market gardeners who are scattered through three or four sections of the city.

The Agricultural Survey was essentially an inquiry among resident farm families, supplemented by a village survey described in Appendix B, and by the listing of significant prices in market towns.

1. Field Procedure.

The family investigators in the City Survey in Nanking were instructed to fill out a family schedule for every family in every 50th inhabited house.

A "House" was defined as a "house number" even though in some instances there were several apartments or buildings at one number.

In March many entrances were barred, and there was some little difficulty in determining which houses were inhabited.

Consequently, some may have been passed over.

A control map served to check areas skipped. Each man was assigned a specified section on the map to cover and to record the count of house numbers in selecting his 50th inhabited house.

The investigators' were well received because of the favorable reputation of the Committee but were careful to explain that they came only to inquire about facts and not as the family relief investigators working in the regular service of the Committee.

Men who participated in both activities thought very definitely that the Family Survey was much less subject to exaggeration of losses than the relief investigations.

The building investigators in the City Survey had two tasks: (1) to count every building in the city and indicate whether it had been damaged by military operations, fire or looting; and (2) to make an estimate of the loss on every 10th building.

For the purpose a house number was considered a "building", though in some cases it included more than one structure.

A well-trained construction engineer worked out unit-cost figures for each of the common types of construction, which greatly facilitated the accuracy of these estimates.

Furthermore, of each pair of investigators, one was a contractor. Estimates on loss of contents in uninhabited buildings had to be based on the nature of the building and inquiries from neighbors.

A control map located areas overlooked and these were carefully re-done.

This "Introduction" is written to inform general readers as to how the surveys were conducted.

Those interested in technical considerations are referred to Appendix A, "Further Notes on Organization and Method."

― 1 ―

Page 2

Both the family and building investigations covered the whole of the city inside the walls and the areas just outside some of the gates as well.

But they did not take in all that formerly comprised the Nanking Municipality, which included Pukow and a number of surrounding small towns.

Only certain small areas and scattered individual houses inhabited by the Japanese military or civilians could not be studied.

In the Agricultural Survey two investigators were sent out across each of the six hsien armed with passes from three organizations.

They were instructed to follow a main road one way and then to return zigzagging across the main road in the form of a figure eight in order to cover the areas back from the main roads.

On this circuit they were to secure a village schedule from every third village on their route and in that village fill out one farm schedule for every tenth farm family that had returned to the village.

A market town price schedule was to be answered for every market town passed through.

2. Time of Data. The field work for the Agricultural Survey extended from March 8th to 23rd.

For the City Survey the family investigation extended from March 9th to April 2nd with supplementary work April 19th to 23rd; the building investigation, from March 15th to June 15th.

During the longer period for the building investigation there was little change in contents already lost, but there were some cases in which partial taking away of building materials occurred.

Re-building during the period was practically nil.

3. Statistical Procedure. Both in training the investigators and in supervising the tabulation work it was fortunate that an experienced man trained in Agricultural Economics was available as supervisor; and some men who formerly worked in Professor Buck's tabulation department were here to help in that work.

In the writing of the report and the interpretation of the results of the survey the Director has had the invaluable cooperation of Dr. M. S. Bates of the University of Nanking whose extensive knowledge of economic history and of conditions in China has thrown additional light on the statistical findings.

Both the Agricultural and Family Surveys were based on a sample instead of attempting complete enumeration.

Consequently, totals and grand averages are estimates based on results learned in the cases studied.

But with the exception of rice seed for Luho, explained in the proper place, the data on which the estimates in the tables are based remain as reported by the investigators.

In the Agricultural Survey the average per farm family studied was worked out by hsien and then that average was multiplied by the total number of farm families in each hsien, as given by Professor Buck in his Land Utilization in China.

The grand total was secured by adding up the hsien totals and any grand averages were then computed from these totals, thereby giving grand averages weighted according to the number of farm families in each hsien.

Although a village schedule was used to get a broader picture of general conditions, all calculations were based on the farm schedule. (See Appendix B.)

The totals in the family investigation of the City Survey were calculated by multiplying by 50 the average per family secured from a study of every 50th

― 2 ―

Page 3

inhabited house.

Likewise the estimates of loss in the building investigation were computed by multiplying by 10 the total loss secured from a study of every 10th building.

In the printed tables, the convenience of the reader has been considered by dropping as many decimal places as possible.

All totals are given in round hundreds.

4. Units of Weight, Measurement and Currency: The measure of grains and vegetables was a shih(market) tan by weight which is 100 shih(market) chin, or one-half the metric quintal.

This is 50 kilograms or 110.23 pounds, which is very close to the English hundredweight (112 lbs.), and is 0.83 of a picul.

The mow used for measure of area is the local mow as reported by the farmers; but in calculations the Kiangning mow has been used, as it represents two-fifths of the cultivated area studied.

It is equal to 0.06067 hectares.

The shih(market) mow (standard mow) referred to on occasion is slightly larger, 0.06667 hectares or one-sixth of an acre.

All monetary values in these reports are in Chinese currency.

For the period covered in the surveys, the Chinese dollar was practically steady at about $3.40 per United States dollar or $17.00 per British pound.

Lewis S. C. Smythe.

― 3 ―

Page 4 I. CITY SERVEY

I. CITY SURVEY

1.POPOLATION

The city of Nanking had before the war a population of just 1,000,000, which was considerably reduced by repeated bombings and latterly by approaching attack and the removal of all Chinese governmental organs.

At the time the city fell (December 12-13), its population was between 200,000 and 250,000.

The persons reported in our sampling survey in March, multiplied by 50, give 221,150 as the population directly represented by the City Survey.

This number was probably 80 to 90 per cent of the total residents at that time, some of whom lived in places not accessible to investigators. (See Note to Table 1 for fuller discussion of population).

27,500 persons were living in refugee camps maintained by the International Committee, 12 per cent of the surveyed population.(1) Outside the camps, but within the Safety Zone Area were an additional 68,000 persons, 31 per cent of the total.

Some idea of the crowding, the price willingly paid for partial security, is indicated by the fact that 43 per cent of the population, 14 weeks after the fall of the city, was living in an area which had only 4 per cent of the total number of buildings noted in the Survey, and which comprised roughly one-eighth of the total area within the walls.

'The fact that practically no burning occurred within the Zone was a further advantage, and suggests the generally preferential treatment given to the Zone area as compared with the destruction and violence outside, even though the Zone was not officially recognized by the Japanese authorities.(2)

(1). At its maximum in the second half of December and in January, this number was 70,000.

The reduction was irregularly due to the following causes: crowding and discomforts of the camps, though they were generally preferred to the greater dangers and difficulties outside; the need to care for homes and remains of property, whenever there was sufficient security to make the effort worth while; encouragement by the International Committee for return to other parts of the city in every possible case; threats of forcible eviction from the camps as of February 4, fortunately not realized in action, but bringing about much unnecessary suffering and many regrettable incidents.

(2). We must here trace the divisions of the city as used for the purpose of the Survey.

The Safety Zone Area was bounded by Han Chung Road on the South; by North Chungshan Road from Hsin Chieh K'ou past the Drum Tower to Shansi Road, on the east and northeast; by Shansi Road on the north, continued to Sikang Road, which formed the westerly boundary.

Within the Safety Zone Area were the Refugee Camps, reported separately.

South of the Safety Zone Area lay Cheng Hsi, reaching to Shengchow Road, and bounded on the east by Chungcheng and Chungliwa Roads. The remaining southwest corner of the city was called Men Hsi.

The southeastern corner, north to Pai-hsia Road and east to Tungchimen, was considered as Men Tung.

The region from Chungcheng Road eastward to the wall was named Cheng Tung. All the remaining northwestern, northern, and northeastern (as far south as East Chungshan Road) sections were considered Cheng Pei.

For purposes of the building investigation, Cheng Pei Tung was split off from Cheng Pei; this eastern section of the north City ran from North Chungshan Road to the wall bounded on the north by the Drum Tower and Peichiko.

The four sections outside the wall are easily recognized by their names; the Shuihsimen Area extended northward past Hanehungmen. (Sec City Survey map of Nanking.)

The family investigators found the area outside Tungchimen deserted.

Con-

― 4 ―

Page 5

The average size of family for all sections was 4.7. Outside the wall, the average was 4.0, suggesting the presence of more men without families, or of more broken families. Compare the 1932 figures for 2,027 families in the same portions of Nanking from which many of the present population are derived.

They show an average family of 4.34.(1)

It is probable that in normal times there are more persons detached from their families for reasons of employment. The data on population are given in Table 1.

DISTRIBUTION BY SEX AND AGE

The March population of Nanking showed clearly the characteristics of a war time population. The present survey reports for all ages, in all sections of the city, a sex ratio of 103.4 (males, to 100 females);(2)

while the 1932 study showed for all ages 114.5; and in the entire population before the war, there was a very high ratio of males to females, at one time 150. The drop of 9 points in the sex ratio since 1932 is accounted for in part by the withdrawal of males not native to Nanking but formerly working here, and in part by the killings of males in the critical period. Most serious is the acute fall in the ratio for the age-group 15 to 49 years, which roughly represents the productive life of the population; here the decline was from 124 to 111, or 11 per cent.

This change presents the fact that a large number of women and children are deprived of men who were the support of the family.

If the comparison is carried into narrower age-groups, fluctuations are found because of the fewer cases forming the base for each figure; but the results for the 25 years of young maturity are sufficiently consistent to be startling: 15-19 years, 108 now as against 123 in 1932; 20-24 years, 106/124; 25-29 years, 100/128; 30-34 years, 89/123; 35-39 years, 105/123. The decline in the males of productive age is shown in another manner.

Of all males in 1932, those 15 to 49 years old were 57 per cent; in the present survey, they were only 49 per cent, a decrease at the rate of 14 per cent, which constitutes a serious economic and social problem. Correspondingly, of all males those over 50 years of age gained from 13 per cent in 1932 to 18 per cent now, an advance of some 30 per cent.

The variations in sex ratios by sections of the city are of some importance. Although the ratio for all sections was 103, for the refugee camps it was only 80, since they were overcrowded with women seeking security; on

sequently, it does not occur in the family investigation but is included in the building investigation.

The normally crowded sections in the southerly portions of the city (Cheng Hsi, Men Hai, Men Tung), were the first to show a fair degree of recovery from the practically complete depopulation of the critical period.

Together they had 81,000 residents, 37 per cent of the total. (By June this number of residents had doubled, according to the City Government records of registration.)

The sections thus far named had practically 80 per cent of the total for the city.

There were only 8,550 persons in the districts studied outside the wall, which suffered so terribly from the burning by the Chinese army and from violence, and in March were still more dangerous on the whole than the inner city.

(1). Smythe, "The Composition of the Chinese Family," Nanking Journal, University of Nanking, November, 1935, v. 5, No. 2, p. 371-393.

(2), The May 31 registration figures of the Municipal Government, patently incomplete for females, show 109.4.

― 5 ―

Page 6

the other hand, in the less secure areas the men were relatively much more numerous, as is shown by the ratios for Cheng Pei, 121, the garden group, 150, outside the wall, 144.

If we consider the ages for which security was a most acute problem, 15 to 39 years, we find in the refugee camps sex ratios running very low, from 40 to 67 in different five-year units; for the Safety Zone Area, roughly 90; for Cheng Hsi over 150, for outside the wall, well over 200.

Thus men were returning first to the more dangerous localities, with old women and children following along more or less closely; but many of the young women remained in places of relative safety.

The data on sex and age are given in Table 2.

FAMILY COMPOSITION

'The families remaining in Nanking were classified as "Normal," that is, either husband and wife or husband and wife with children living together; "Broken," man or women with children; and "Non-family," man alone or woman alone.

'Then each of these three types was repeated "with relatives." 'The "normal" families were much fewer in proportion to all families than in an earlier study made among Nanking people in more settled times, 1932: now only 4.4 per cent with husband and wife as compared with 9.5; now only 26.2 per cent with husband, wife and children instead of 33.1 per cent.

This represents a reduction of these types by one-fourth.

A slight increase in "Normal with Relatives" occurred: 32.3 per cent as compared with 29.8 per cent in 1932.

In other words there is a net loss of normal families amounting to 9.5 per cent of the total families, or one-seventh reduction of normal families.

This decline in normal families is largely due to an increase in the broken families, 21.4 per cent as compared with only 12.9 in 1932, or an increase of 8.5 per cent for the four types of broken families. Of that increase 6.9 per cent was in families without a man for support, that is, families consisting of only women with children.

This means that the number of broken families was almost doubled. This increase in broken families is more clearly understood when we realize that 14.3 per cent of the members of the families remaining in Nanking had migrated, but only 2.2 per cent of the wives lost husbands by this migration.

In addition to these there were 4,400 wives, or 8.9 per cent of the wives, whose husbands had either been killed, injured or taken away. Two-thirds of these were killed or taken away, 6.5 per cent. Or more poignantly, 3,250 children (5 per cent of all children) had their fathers killed, injured or taken away.

'These broken families could only to a small extent be due to families being divided within the city because only 3 per cent were so reported.

The three factors combined of migration, persons killed or taken away, and divided families, broke 11.7 per cent of the families remaining in Nanking, or 5,500 families.

Within the city the refugee camps showed a very high figure for broken families, especially in the case of women with children, 13.2 per cent as compared with 6.6 per cent for all sections and with 3.4 per cent in the more normal times of 1932.

Fourteen per cent of the families in the refugee camps were women, children and relatives (the latter usually dependent).

Altogether, 27.2 per cent of the families in the refugee camps were women with children and in some cases with dependent relatives.

In the camps 35 per cent of the

― 6 ―

Page 7

families had a woman as head, while in the remainder of the population only 17 per cent of the families had a woman head.

A man or woman alone constituted the family in 14 per cent of the families living outside of the wall, as compared with general figure of 7.4 in 1932.

In another 16.3 per cent of the families outside the wall, the family was a man with relatives.

For family composition analysis see Table 3.

2. DEATHS AND INJURIES DUE TO HOSTILITIES

Number and Cause

The figures here reported are for civilians, with the very slight possibility of the inclusion of a few scattered soldiers.

The reports made in the Survey indicate that 3,250 were killed by military action under known circumstances.

Of those killed 2,400 (74 per cent) were killed by soldiers' violence apart from military operations.(1)

There is reason to expect under-reporting of deaths and violence at the hands of the Japanese soldiers, because of the fear of retaliation from the army of occupation.

Indeed, under-reporting is clearly emphasized by the failure to record any violent deaths of young children, although not a few are known to have occurred.

Of the 3,100 receiving injuries under known circumstances, 3,050 (98 per cent) were definitely by soldiers' violence aside from warfare.

There was a noticeable tendency to ignore injuries from which some sort of recovery had been made.(2)

89 per cent of the deaths and 90 per cent of the injuries by soldiers' violence occurred after December 13, when occupation of the city was entirely completed.

In addition to those reported killed and injured, 4,200 were taken away under military arrest. Persons seized for temporary carrying or other military labor were seldom so reported.

Very few of those here mentioned were heard from in any way up to June. The fate of others gives reason to think that most of them were killed early in the period.(3)

The figures for persons taken away are undoubtedly incomplete.

Indeed, upon the original survey schedules, they were written in under the heading "Circumstances," within the topic of deaths and injuries; and were not called for or expected in the planning of the Survey.

Thus they have an unusual

(1). "Military operations" is here used for shells, bombs, or bullets fired in battle.

(2). Among the injuries reported to our Rehabilitation Commission by the 13,530 families applicants for relief, whom they investigated during March, was rape to the extent of 8 per cent of all females of 16-50 years.

This figure is a serious under-statement, since most women who suffered such treatment would not volunteer the information, nor would their male relatives. While raping was such a common matter in December and January, people were much freer in admitting rape than under ordinary circumstances. But by March families were trying to hush up the fact that women in their families had been raped.

The matter is mentioned here as illustrating the acute insecurity from which the social and economic life of the city has suffered.

(3).The seriousness of "taking away" is underlined by the fact that all so listed are males. Actually many women were taken for shorter or longer service as waitresses, for laundry work, and as prostitutes. But not one of them is listed.

― 7 ―

Page 8

significance, and are more important than the simple figures indicate. Thus, those 4,200 must contribute an important addition to the number killed by soldiers.(1)

Ignoring many minor cases, the casualties of military operations and the sum of those killed and injured by soldiers' violence, and of those taken away, represents 1 person in 23, or 1 in every 5 families.

The critical social and economic results of these killings are in part indicated by the following direct calculations from our listings. The number of women whose husbands were killed, injured, or taken away, was 4,400.(2)

The number of children whose fathers were killed, injured, or taken away, was 3,250.

Of the 6,750 persons violently killed and injured, only 900 (or 13 per cent) came to grief through military operations.

The data on numbers of deaths and injuries are given in Table 4.

DISTRIBUTION BY SEX AND AGE

If we analyze by sex and age those who suffered violence or abduction, we find that the per cent of males in the killed and injured was for all ages 64, and reached for ages 30-44 the high percentage of 76.

Able-bodied men were under suspicion of being ex-soldiers; many were killed for having callouses on their hands, supposed evidence of carrying rifles.