女子労働

論文一覧に戻る
女子労働

Technology Change and Female Labour in Japan

論文タイトル: Chap. 3: Female Workers of the Urban Lower Class
著者名: Miyake, Akimasa
出版社: United Nations University Press
出版年: 1994
本論文の目次を見る本ページのPDF版を見る

Chap. 3: Female Workers of the Urban Lower Class


This chapter will examine the state of urban lower-class female workers mainly in the period between 1870 and 1920. Although earlier academic works have not focused specifically on female labour as such, much research has been conducted on the pre-war urban lower class. Tsuda Masumi's indepth 1911-1912 survey of "indigents" (saimin) emphasized the proletarian nature of pre-war Japanese society at large.1 Sumiya Mikio saw the stratum of urban lower class employed in miscellaneous occupations as a pool of surplus labour and the basis for the supply of wage labour.2 Hyodo Tsutomu traced the process by which large-scale factory workers rose out of the urban lower class between the 1880s and the post-First World War period.3
While the conclusions drawn in these studies vary, all their authors address the urban lower-class issue only in terms of the formation of the wage labour stratum. Regardless of whether the emphasis is placed on the decline into the lower class or the rise out of it, virtually all interest is directed at the formation of the "wage labour" sector. This approach seems to be justified, considering that the development of capitalism leads to class stratification and necessitates the creation of wage labour.
However, while this approach is acceptable from a theoretical standpoint, in reality the urban lower-class condition was not wholly a reduced state from which wage labour attempted to escape. The urban lower class underwent internal change while concomitantly developing (or perhaps retrogressing) independently during the development of capitalism.
Research by Nakagawa Kiyoshi develops along comparable lines. He approaches the pre-Second World War urban lower class, especially that in the inter-war period, as a dynamic element of urban society, and he delineates the changes in the living conditions of this group.4
This chapter, which owes much to Nakagawa's research, focuses on the role of female labour as a part of the urban lower class during the 50 years between 1870 and the 1910s.5

I. 1870 to the 1880s

1. Historical Antecedents

The historical antecedents to the development of the urban lower class during the rise of capitalism can be found in the status of the museki musan no kyumin (the destitute who possessed neither family registers nor assets of any kind) of Edo at the end of the feudal period. By the eighteenth and nineteenth centuries, the city of Edo had become a huge urban centre of one million people. Unfortunately, few demographic records on the living areas, living standards, occupations, and family structure of the destitute remain, and existing fragmentary references must be pieced together.
The Fujiokaya nikki6 records the living areas of 74,000 "destitute persons" to whom the town councils sold rice at exceptionally low prices in 1860 in an effort to defuse the upsurge of rice uprisings. According to this source, the destitute lived in a wide area encompassing Yotsuya, Shitaya, Shiba, Fukagawa, Honjo, Koishikawa, Kanda, and Hongo, including Yotsuya Samegahashi, Shitaya Mannencho, and Shiba Shin'ami. In the 1890s, the last three areas became notorious as Tokyo's three largest slums.7
As seen below, from the seventeenth and eighteenth centuries these areas had become the ghettoes of the poor. Sources note that many beggars gathered in the lower-lying areas of the Yotsuya Samegahashi, a former swamp site which had been reclaimed in the early Edo period. Shiba Shin'ami first became a settlement for beggars and lowly entertainers in the late seventeenth century. Shitaya Mannencho, which was formerly called Yamazakicho, gradually grew into a lower-class community from the tenement houses built by masterless samurai Yamamoto Nidayu for ragpickers and other lower-class individuals.8
The development of the commodity economy in the late eighteenth and early nineteenth century accelerated the breakdown of the agricultural stratum, and resulted in a build-up of penniless people and destitute farmers who were forced to leave their villages. Most headed for large cities such as Osaka and Edo and settled in the ghettos noted above.
Using the 1828 Machikata kakiage [Records of the Commoner Population], Matsumoto Shiro makes the following points about the lower classes' living conditions.9 First he categorizes the types of households noted in Machikata kakiage into those of home-owners, landowners, landlords, property tenants, house tenants and akidana (vacant shops). If house tenants are considered the standard for determining the dividing line for the lower class, then a high proportion of the population fell into the lower class, as this sector accounted for over 60 per cent, on average, of all households in each ward of Edo. "Destitute persons" must be considered only a small proportion of the lower class.
As for actual living space, a popular contemporary phrase expresses it simply―"the back alley tenements measured 9 shaku 2 ken," which is equal to approximately ten square metres. Naturally all well and lavatory facilities were shared. These tenements were located in dark and dreary back alleys, and residents had to make the best of the cramped space and dearth of sunlight.
While the descriptions above shed some light on the living area and living conditions of the lower class, it is more difficult to determine the jobs performed by them. Matsumoto investigates this problem using, the slightly later historical records dating from the period when the newly founded Meiji government and big merchants planned to banish the lower-class elements involved in the rice uprisings to the wilderness Shimosa Plain in Chiba, under the rubric of a "pioneering conquest." Although fragmentary, this source provides an insight into the occupations of the lower class, as seen in table 3.1. On the basis of this table, the lower class can be categorized into four main groups.
― Physical labourers: This group made up the largest proportion of workers of all four categories. The majority were cart-pullers, cart porters and day labourers. Cart-pullers and cart porters were involved in the transportation of commodities, while day labour generally referred to all forms of physical labour.
― Craftsmen: This category included those who worked at customers' residences, such as plasterers and carpenters, and those who worked in their own workshops, such as tatami mat makers and cask-makers. Although the fully-fledged dissolution and decline of this sector did not occur until the latter half of the first decade of the 1900s, many such craftsmen were reduced to the lower class earlier,10 since these types of professions were easily affected by changing economic conditions.
― Small-scale merchants, fishermen, grocers: Fishmongers and grocers were primarily itinerant vendors. According to Yoshida Kyoichi,11 who uses different historical references from Matsumoto to study the lower class of the eighteenth and early nineteenth centuries, many women worked by selling anise and flowers (for religious offerings).
― Farmers: Proportionally, farmers made up a small percentage of the lower class. They worked extremely small fields on the outskirts of Edo.
While not included in table 3.1, other historical sources cite a fifth occupational category, or miscellaneous occupations. This group was made up of a jumble of occupations―from itinerant entertainers, diviners, masseurs, and theatre ticket takers to beggars.12 Women were prevalent in such side work as clothes-washing and errand-running.
No historical references remain to help us ascertain whether the members of the lower class formed families or not. Probably they did, to some extent, especially craftsmen, and the employment of wife and children must have been the norm.
Table 3.1a. Former Occupations and Residences of Farmers of Niwamura, 1869
A new breed of lower-class poor appeared immediately after the Meiji Restoration (1868). Large numbers of lower-class samurai, who had been stripped of their privileges, were reduced to poverty. The national and prefectural governments instituted various measures aimed at aiding this group of ruined samurai. In 1869, the Tokyo Metropolitan Government established two relief centres where the family members of ex-samurai were given occupational training. The children were taught weaving, as well as techniques for making silk thread, fabric clog thongs, Indian ink, and paper.
The governor of Tokyo also called together the city's influential businessmen to form a building and repair chamber of commerce. This organization opened a workhouse in 1873. As of 1876, the workhouse came under the direct control of the Tokyo government and began the manufacture of leather shoes, paper, and matchboxes.
An institution for the poor was also established by the Osaka government in 1871. The next year it was reorganized into an occupational training centre, where weaving, papermaking, and spinning were taught. The Kyoto government also set up a similar establishment in 1870. Here too, oil-pressing, candle-making, paper manufacture, weaving, and comparable skills were taught. In 1875, an establishment was inaugurated in Kyoto in which the destitute were paid allowances for collecting garbage.13
Table 3.1b. Former Occupations and Residences of Farmers from Hatsutomimura, 1872
In 1874 the national government promulgated a relief regulation which enlisted the magnanimity of the general population in assisting the lower class, while minimizing national aid as much as possible. As a result, the policies noted above, instituted at the prefectural and metropolitan level, represented the relief measures in real terms. In 1882-1883, however, these policies were either abolished or left to be implemented by private organizations. Commoners as a whole became much poorer as a result of Finance Minister Matsukata's deflationary policies, so that providing relief to only a part of this class was meaningless.14

2. The Expansion of the Lower Class

The so-called "primitive accumulation" stage of capitalism in Japan (in which labour supply, monetary resources, and production methods had accumulated to levels sufficient to form the basis of capitalistic development) can be dated to the period between the Matsukata deflationary policies and the earlier part of the 1890s. It has been said that, unlike in Western Europe, where large masses of poor suddenly appeared in the urban centres, the numbers of persons displaced from the countryside to the cities were relatively small during this period of primitive accumulation.
In recent research, Unno Fukuju contests this theory. He insists that labour statistics point to a large inflow of farmers into the cities. In fact, of the 920,000 persons residing in cheap lodging-houses in Tokyo in 1884, only 44 per cent were registered as Tokyo residents, while the remainder had come from other areas throughout the nation.15
Historical records state that the 10,000 residents of the Nagomachi area of Osaka, one of Japan's four largest slums (the other three were Yotsuya Samegahashi, Shitaya Mannencho, and Shiba Shin'ami), were divided into two groups: the original residents, who had fallen into dire circumstances, and the newly settled residents.16 There was also a considerable increase in the inflow of seasonal labour into Tokyo at this time.17
Owing to the rapidly growing numbers of the poverty-stricken, the settlements for the poor began to expand from the original ghettoes to the surrounding areas. The residential status of the expanded lower class is not thought to have changed from earlier patterns.18 In this way, the urban lower class became inflated through the "primitive accumulation" process.
Eventually the rapid growth of the lower class attracted wide social attention. Surveys on slum conditions were taken up in several newspapers. Of these, the following two mentioned the state of women workers in depth.
Kure Ayatoshi's survey on the status of Tokyo's destitute, published in volume 57 (1891) of Sutachisuchikku zasshi [Statistics Magazine],19 was "a collection of excerpts taken from a certain newspaper's survey of the state of Tokyo's poor during the escalation of rice prices last year." In it, the occupations of men and women from each ward of Tokyo ware noted as follows.

Day labourers, construction workers, jinriki-pullers, merchants at temple fairs, low-grade factory workers, used paper and junk collectors, clog-repairers, dancers, outdoor storytellers, outdoor exhibitioners, fishermen, shell fishermen, bait-diggers, peddlers, papermakers, makers of paper ornaments for hair and decorative uses, fishmongers, and vegetable-sellers. For women and children, cart-pushing and itinerant tea-picking.

While "low-grade" factory workers are not discussed in detail, this is taken up in Matsuhara Iwagoro's 1893 work entitled Tokyo shinai no shukojin kato shakai seigyo no ichidai gensho [The Serious Situation of Lower-level Manual Labourers of Tokyo].20 These people could be considered as the manual labour sector of the miscellaneous occupations group. For example, many women and children worked in factories which manufactured umbrella ribs in what was once side work but developed into a primary source of work. In the Honjo area of Tokyo alone, there were 8,000 of these manual labourers, and of them two-thirds were women and children. During this period, most lower-class women workers did not find themselves in circumstances greatly different from those of the late Edo period. However, the appearance of new types of low-grade work merits attention.
Suzuki Umeshiro's 1888 survey of the Nagomachi slum in Osaka, Osaka Nagomachi hinminkutsu shisatsuki, published in 1918,21 gives a specific breakdown of the occupations of those living in Nagomachi. These occupations have been categorized in table 3.2. The following points concerning female labour can be drawn from the table. First, one of the first industries to appear during the early period of capitalism was match-manufacturing, which employed a relatively large number of workers. In sharp contrast to the small percentage of men, over 10 per cent of all women were engaged in this work. The actual working conditions in the match-manufacturing industry will be taken up in the next section.
Secondly, in terms of occupation, the largest proportion of persons (16.4 per cent of all workers) were involved in garbage-collection. With the inclusion of miscellaneous jobs, such as old clog collection and sales of used casks (which fall in the category of "miscellaneous jobs" in the table), the percentage of persons involved in the collection of discarded objects amounted to close to 30 per cent.22 Both men and women were involved in this work in approximately equal proportions.
Thirdly, approximately 26 per cent of the women were unemployed. While it is natural for girls under 15 to be unemployed, it is notable that a fair proportion those over 15 were not employed. It is doubtful, however, that these women were entirely without work. Suzuki notes that theft was another type of occupation among the poor of Nagomachi. Taking into account this type of "occupation," the numbers of so-called unemployed are reduced by a large margin.
Table 3.2. Work Performed by the Poor of Nagomachi, 1888
Based on the records from Nagomachi and Tokyo, it is possible to categorize as follows urban lower-class female labour in the 1870s and 1880s.
― Physical labour: This category applies to tea-pickers or cart-pushers in Tokyo, and day labourers in Nagomachi. As in the late feudal period, this group encompassed many different kinds of people engaged in physical labour.
― Craftsmen: this included fan-makers, rope-makers, rope-weavers, and other labour involved in manual production. It is likely that most of these labourers did such work as a side job at home.
― Miscellaneous occupations: This included such work as rubbish-collecting, entertainment, and petty trading operations.
― Factory work: The leading type of factory work was related to match-manufacturing, one of the first occupations to utilize lower-class women during the process of capital formation.
These four categories were in no way fixed in terms of substance or classification. While the respective occupations had a certain degree of independence from each other, there was also a degree of fluidity among the occupations―an entertainer could turn to rubbish-collecting, for example.23 During this time, occupational trends turned away from all other categories toward factory work. Little is known about the state of lower-class families during the 1870s and 1880s. Judging from the age brackets found in table 3.2, it is certain that lower-class workers had families. However, since the head of the household and the spouse had extremely meagre incomes, it is likely that all family members were employed in some capacity.

II. The Industrial Revolution

1. The Development of the Urban Lower Class

Japan's industrial revolution began in 1886-1889, with the sudden rise of private enterprise, and went on to 1907.24 As shown in tables 3.3 and 3.4, Tokyo's slums expanded greatly during this period. Table 3.4 shows the number of cheap lodgings in each ward. These cheap lodgings housed the displaced rural folk who flocked from throughout the country to the cities in search of work. The tables demonstrate how the areas where the lower class congregated differed considerably before and after 1896. From the period of "primitive accumulation" to the early industrial revolution period, most lived in the Asakusa-Shitaya areas, whereas from the mid to late industrial revolution era, the highest concentrations of poor were in the Fukagawa and Honjo areas, which were Tokyo's industrial centres at the time. This trend reflects the growing numbers of the lower class involved in factory work.
Table 3.3. Appearance of New Slums in Tokyo
Table 3.4. Numbers of Cheap Lodgings in Tokyo, by Ward
Table 3.5 shows the numbers of private factories in Tokyo from 1889 to 1907. From this period to the early industrial revolution period, the factories were primarily involved in the production of matchboxes, rubber, and soap, as well as various types of chemical industries. As can be seen from the table, textile and heavy industry factories increased gradually. Osaka, which was then Japan's largest industrial centre, saw a comparable trend.25
Table 3.6 tabulates the changes in the number of female workers in Tokyo, by occupation. Both tables 3.5 and 3.6 clearly reflect the high proportion of labour in the chemical industries (over 90 per cent were involved in the match industry) at the start of the industrial revolution, and the shift to the textile industry, particularly the spinning industry, in the 1900s. (The reason for the low percentage of workers in table 3.6 in the heavy industries, which had many factories in the latter part of the industrial revolution―see table 3.5―is the low percentage of female labour in these fields.)
All these industries turned to urban lower-class women as a source of labour.26 The discussion below considers the match industry as a casestudy.27
Table 3.5. Changes in the Number of Private Factories in Tokyo


2. Female Labour in the Match Industry

Match-manufacturing technology, which was introduced to Japan during the 1870s, quickly became established due to the simplicity of the process, and soon became one of Japan's leading export industries. According to a study by Takamura Naosuke on the industrial and trade structure during the industrial revolution,28 from 1898 to 1900 matches accounted for 3.1 per cent of Japan's average annual export volume, and were ranked fourth behind raw silk thread, cotton thread, and silk fabric (habutae). By the latter part of the industrial revolution, this figure had shrunk to 2.5 per cent, and by 1910 matches were ranked seventh or eighth among Japan's major exports.29 As seen from table 3.7, over 80 per cent of the matches produced were for foreign consumption. In the early phases of match production, the main importing country was China; later, importing countries ranged across South-East and South Asia.
Table 3.6. Fluctuations in Female Occupations in Tokyo
By the industrial revolution, matches were produced throughout all regions in Japan, although in 1895 90 per cent of production was concentrated in Hyogo, Osaka, Aichi, and Tokyo. As for the concentration of factories, in 1899 the above-mentioned four regions accounted for 70 per cent of all factories (with more than 10 employees) nationwide, with 62 factories in Hyogo, 35 in Osaka, 12 in Aichi, and 17 in Tokyo. These four regions also accounted for close to 80 per cent of all the employees in the match-manufacturing industry.30 Hyogo Prefecture and Osaka boasted the highest figures in terms of production volume, factory numbers, and numbers of employees. This can be accounted for by the proximity of Kobe, the leading trading port at the time; this facilitated the transportation of matches, which had become an export commodity.
Table 3.7. Match Factory Indices
Match manufacturing was conducted on a medium-sized to tiny scale. According to the 1909 Statistical Records on Factories, there were no match factories with over 1,000 employees. Factories employing between 5 and 99 workers accounted for over 70 per cent of all match-manufacturing factories. This situation remained essentially unchanged. The simple nature of the match-manufacturing process accounts for the small scale of the factories and the wide employment of female labour, which formed the majority of urban underclass workers.
The main processes involved in the match-manufacturing process were three: production of the matchsticks, and construction of the matchboxes and of the match head itself. Most match factories were supplied with matchsticks and matchboxes produced elsewhere.31 Match splint manufacturing plants were initially centred in the Hokkaido and Tohoku areas, owing to the proximity of white poplar forests, which provided the raw material, but in the 1900s they began appearing in the Kobe-Osaka area, where match manufacturers were concentrated.
As for the matchboxes, generally the match manufacturers would purchase the materials and farm out the assembly work and work of applying the brandname seals to the boxes. This work was usually farmed out to urban lower-class women, who did it at home as piecework. Case-studies from the Kobe-Osaka area reveal that these women were not directly hired by the match manufacturers, but by the small matchbox-makers working on contract from the manufacturers.
Excluding the matchstick and matchbox production process, the manufacture of matches involved the following ten steps: preparation of the wood splints, alignment of splints, stamping, application of paraffin, application of the chemical composition, drying, packing into boxes, cleaning of the striking surface, final inspection, and packaging. Of these, the most technically important aspect was the alignment of the splints for dipping the match head.
As shown in table 3.7, women accounted for 70 per cent of match-manufacturing factory workers. Most of them were involved in three of the above operations―the alignment of matches, packing, and packaging―as noted in the Ministry of Agriculture and Commerce report on the conditions of factory workers, which states that "the primary job of women in the match-manufacturing industry was preparing the matchsticks (although boys also did this type of work). Packing the matchboxes and wrapping the packages were done solely by women."32 Female workers were pivotal to the match-manufacturing industry, both quantitatively and qualitatively.
It would be expected that the technological revolution of the matchstick alignment process, the most technically demanding aspect of the match-manufacturing process, would reduce the numbers of women workers. In the early 1890s, a German matchstick-alignment machine was introduced. Because "the German matchstick-alignment machines can only be operated by adult men,"33 it was assumed that women would be excluded from the alignment process. However, very few factories were able to introduce the German machinery, owing to its high cost. Instead, in the late 1890s, a Japanese-made foot-pedal-operated alignment machine was introduced and its use began to spread. These machines were operated by older women; thus, women were not eliminated from the operations even after automation with the Japanese equipment.34
The introduction of this equipment helped raise the productivity of Japanese match factories. As seen in the right-hand column of table 3.7, per capita productivity increased 1.5-fold between 1905 and 1906, a period that coincides precisely with the introduction of the Japanese matchstick-alignment machines.
Table 3.8. Case-study of Match-factory Workers and Their Familiesa
Regarding the supply of female labour in the match factories, Yokoyama Gennosuke notes on page 126 of Nihon no kaso shakai:

Regarding the relationship between match factories and the lower class, in places such as Kobe, where maintaining a livelihood is uncertain, the growth of slums can be attributed to multiple causes. In areas such as Osaka, however, where there are few occupations for the lower class, many people are reduced to poverty and life in the ghettoes. While the change in the appearance of Nagomachi, Japan's leading slum, can in part be attributed to other factors, the leading element behind this change is the establishment of many match factories near the area.

As this commentary suggests, the match factories were first established and underwent development based on a supply of urban lower-class labour, particularly young workers. Moreover, the appearance of the match factories helped to alter the living conditions of the urban lower class. Page 153 of Matchi shokko jijo [Conditions of Match-factory Workers] describes the relation between the match factories and the urban lower class in the following manner:

Owing to the fact that match-factory workers are generally the poor from the ghettoes, the two are bound by an inseparable tie. The factory workers commute from their homes or from cheap lodgings, and most match factories.are located on the outskirts of town, not far from these slum areas. The reason for locating factories in such areas is to ensure a steady supply of factory employees. Typical examples include certain portions of Honjo, Fukagawa, and Asakusa in Tokyo, areas adjacent to Imamiya and Nanba in Osaka, the Tachibanacho-dori, Kawasaki, and Fukiai areas of Kobe, and the outlying areas of Hyogo.

Matchi shokko jijo notes that the best way of determining the living conditions of the match-factory workers is to investigate the extent and types of living expenses incurred by those living in the ghettoes, and provides case examples of the living conditions of 22 urban underclass families in the Osaka area. Of these, 15 families (table 3.8) were connected with the match factories in some capacity. The following four points can be induced from the table:
1. Families with family members employed at the match factories were all multiple-income families. Match-factory workers were generally adult women (mainly wives) or adolescent boys and girls. Most of the family members were forced to work since the income of the head of the household was far from sufficient to raise the entire family.
2. Since several members of each family were generally employed, the wages paid by the match factory were considered supplementary income and, as a result, were extremely low. Those receiving a daily wage of less than 10 sen (one one-hundredth of a yen) greatly outnumbered those with higher incomes. In this connection it should be mentioned that even women employed as power-loom operators, who, according to annual statistics, had the lowest wages among the various job sectors, received approximately 20 sen in daily wages in the first half of the 1900 decade.35 These low wages of match-factory workers kept their family finances chronically debt-ridden, as can be seen from examples 1, 2, 7, and 12 in table 3.8. Yokoyama's remark that the emergence of match factories altered the look of the lower class is relative. Even when the match factories were at their height, the conditions of the underclass did not improve markedly.
3. The assembling and labelling of the matchboxes were tasks primarily farmed out to women who did the work at home. Their income was even less than that of the match-factory workers. Generally a worker assembled 2,000 boxes a day. The wage per 1,000 boxes was 7 sen, with 1 sen 5 rin (1 rin is one-tenth of a sen) deducted for paste, leaving 5 sen 5 rin.36
4. During the two decades of the 1870s and 1880s, most urban underclass families had a family member employed in either the match-manufacturing factories or matchbox assembly work. As seen in table 3.8, match-factory work gradually came to replace physical labour and other factory work as the primary occupation of the head of the household.

3. Factory Conditions

Now let us look at conditions in the match factories. The following sections will introduce the labour composition, employment structure, consecutive years of labour, and working conditions of the match-factory workers.

(1) Composition of the Labour Force

In this section, we will look at the age structure and functions performed by female labour. The age. structure of workers is described in Matchi shokko jijo [Conditions of Match-factory Workers], a report on a study by the Agriculture and Commerce Ministry on 14 medium-sized (from 34 to 1,200 employees) match factories in Osaka between 1900 and 1902 (table 3.9). Minors of under 14 years of age were found by this study to account for 18.9 per cent of the workforce, but this figure is qualified with the statement that: "Although in most factories the minimum hiring age is 12 or 13 years of age, even very young children without any training can earn 2 or 3 sen a day working at a match factory; impoverished children lie about their age to get a job; factories let them work unless circumstances force them to investigate." Thus workers under 14 years of age accounted for a greater percentage than indicated in table 3.9; the cumulative percentage of workers of under 20 years of age indicates that the match-factory workforce was clearly composed primarily of young female workers, that is, girls and young women.
Table 3.9. Age of Workers in Match Factories, 1900-1902
Table 3.10. Division of Roles by Age and Sex among Workers in Match Factories, 1902
Matchi shokko jijo gives detailed information on the work they performed in five factories (table 3.10). Of the different stages of match production, box labelling was usually done outside, so the data in the table refer to only three of the five factories studied. Female workers working in the factory were primarily responsible for filling boxes and lining up matchsticks; the majority of workers doing these jobs were under 20.
In short, girls and young women were the mainstay of the match factory workforce, and most were engaged in labour-intensive manual activities (tables 3.9 and 3.10).

(2) Form and Duration of Employment

Match-factory workers "lived for the most part in nearby poor neighbourhoods and none came from very far away; thus no special recruiter had to be sent out to find workers. Usually, workers were recruited by posting signs at the entrance to the factory or on street corners."37 Unlike heavy industry, where new workers were introduced through the offices of an oyakata (boss), or the spinning and silk-reeling industries, where recruiters supplied female labour, the match industry did not actively recruit, but tended to rely on local hiring because of the inherently small scale and the siting conditions of match factories. The simplicity of this hiring approach led to "the absence of contracts specifying periods of employment or advance payment of wages" and frequently "made the work of a match-maker utterly unplanned and irregular, with workers working one day at one factory and the next day at another occupation or at a different match factory."38 Thus, "needless to say, periods of continuous service were short,"39 and this tendency was especially pronounced among female workers.

(3) Working Conditions

Working hours in most factories were usually "from sunrise to sunset," but in the winter, when the demand for matches was greatest, "work continued uninterrupted until about ten o'clock." On average, therefore, the working day was 14 hours long. Elementary-school pupils "came to work early in the morning and worked for one or two hours before going to school, and returned after school to work until sundown." Apparently, therefore, workers laboured only part of the day in exceptional cases. Workers usually had two days off a month.
Wages were paid on a daily basis in the case of men and on a piecework basis in the case of women. This was related to the nature of the work done: lining up the matchsticks, filling boxes, pasting labels, and wrapping (jobs done mostly by women) were paid on a piecework basis, and applying the match composition to the tips, drying the matches, and mixing the composition (done mostly by men) were paid as daily wages.40 Table 3.11 shows daily earnings at nine factories whose data are reported in Matchi shokko jijo, divided into 10-sen brackets. Match-factory workers were clearly badly paid, especially women and children. In addition to low wages, moreover, working in a match factory involved considerable danger. Fires occurred frequently and each time left many dead and injured, but "not a single factory gave relief to the victims."41
Table 3.11. Wages of Workers in Nine Match Factories, Breakdown by Age and Sex, 1901
The match-making industry relied on poorly paid labour by young women and children of the urban lower class under terrible working conditions. From the viewpoint of the lower strata of urban society, match-factory work turned poor young women and children into an urbanized working class, giving them a role in Japan's capitalist growth process and changing the labour structure from within without dislocating them from their homes in the slums of Japan's cities.

III. The Post-First World War Period

1. The Urban Lower Class after the Russo-Japanese War

Japanese capitalism came to the end of its industrial revolution and began the transition to monopoly capitalism after the Russo-Japanese War of 1904-1905. It was during this period that the Ministry of Home Affairs (Naimusho, Chihokyoku) conducted the first serious studies of the urban lower classes. The first, entitled Saimin chosa tokeihyo [Statistical Tables from a Study of the Poor], was conducted in 1911; the report was released the following year. The second was conducted in 1912 and published in 1914 as Saimin chosa tokeihyo tekiyo [Synopsis of Statistical Tables from a Study of the Poor]. After the Russo-Japanese War, an arson incident at Hibiya in 1905 sparked a nationwide rash of riots by city dwellers, large numbers of them poor.42 The Ministry of Home Affairs' studies were part of a government effort to determine the condition of the poor in the cities and to check discontent before it developed into an organized opposition movement. Using these studies, let us first look at the realities of female labour in the urban lower class around 1910. We will then look at the changes that took place after the end of the First World War, using data on Tokyo to compare the two periods.
The populations of Japan's largest cities climbed sharply after the industrial revolution, mainly due to the sharp increase in the number of people employed by industry. Tokyo's population was just below one million until the mid-1880s; by 1905-1910, it had effectively doubled. Half of its inhabitants had come from other parts of Japan, and their official domiciles (honseki) were not in Tokyo.43
Though the population came to be concentrated in large cities in the course of industrialization, new arrivals did not necessarily begin immediately to work in factories. The majority settled first in urban slums, swelling the ranks of the poor. The exact number of the poor cannot be accurately determined, but data on those officially termed indigent (saimin) in 11 wards of Tokyo are shown in table 3.12. The definition of indigent according to the Tokyo city authorities who compiled these figures is: "coolies, cart-drivers, day labourers, etc., who are unable to pay their ward charges (kuhi) and earn 20 yen a month or less or pay house rent not exceeding 3 yen."44 This category of the poor accounted for 12 per cent of the population.
Table 3.12. Distribution of the "Indigent" in Tokyo, 1911
The indigent were concentrated primarily in wards of Tokyo where, from the 1890s to the first decade of the 1900s, new factories had sprung up and the number of industrial workers was rapidly growing―Honjo, Fukagawa, and Koishikawa―and in the traditional slum areas of Tokyo―Asakusa and Shitaya. These data are therefore in agreement with observations regarding table 3.3 above. Because the increase in the population of the lower class occurred in the industrial zones, it seems justified to interpret the change in the nature of urban lower-class society as having been caused by industrial development. Indeed, the composition of the urban lower class during the period from the Russo-Japanese War to the First World War underwent a change, the most apparent feature of which was the growth in female labour.
Table 3.13 breaks down the occupations of the indigent in four wards of Tokyo on the basis of a 1911-1912 Ministry of Home Affairs survey. Because the wards covered (Shitaya, Asakusa, Honjo, and Fukawa) represent both old and new slum districts, the data give a fair idea of the jobs held by the poor during this period. The preponderance of factory workers is immediately apparent. From 1911 to 1912, the largest category of the poor with jobs, whether heads of households or dependents, male or female, was of those employed in industry. There is no doubt that, compared to the previous period, the predominant occupation of the lower class was now the industrialized type. Women particularly, whether heads of household (mostly widows) or members of male-headed households, were overwhelmingly employed as factory workers.
Table 3.13. Occupations of the "Indigent" in Four Wards of Tokyo
It is interesting to note that among workers from male-headed households, the number of women is over four times as great as the number of male workers (excluding household heads). The reason is that in almost all cases they were the householder's spouse. In this survey, 81.3 per cent of employed women (excluding female household heads) were spouses; 84.5 per cent were at least 20 years of age.45 This speaks for the role of the family and the formation of households in the urban lower class. But for the poor of this period, the "only way to support a small family (the spouses) was to expel from the family at a relatively early age a considerable number of dependents."46 Minors working and living outside (higenjumin, literally non-resident family members) were almost exclusively the children of such families. Upon completion of elementary school at about 12 years of age, children were sent off on their own and began to live outside the family. Though urban poor dwellers began forming and maintaining families, this was made possible by expelling an average of 0.40 "non-resident family members" per household.47
Whether heads of household or not, female job holders were most often employed in spinning or the manufacture of apparel and accessories. The latter occupation was generally engaged in at home, whereas most workers in the spinning industry commuted to factories in the spinning centres of Honjo and Fukagawa. Tokyo spinning-mill owners had just begun to build workers' dormitories and recruit female workers from remote areas, but the urban lower class continued to furnish a certain amount of mill labour.
The primarily industrial employment structure of the poor of this period was not peculiar to Tokyo; it also applied to Osaka. Table 3.14 shows occupations, monthly wages, and days worked by poor female workers in two wards of Tokyo and four districts (machi) of Osaka. In Osaka, industry was not centred on textiles as in Tokyo; the majority of female workers were employed in the chemical industry (mainly match-making) and apparel and accessory manufacture. In Osaka, as in Tokyo, therefore, the main branch of female employment was industrial.
Wages and days worked by women are also shown in table 3.14. In both Tokyo and Osaka, factory workers worked about the same number of days per month or slightly more than in other occupations, but wages (monthly) were lower. Though detailed data on daily working hours are unavailable and a simple comparison cannot be made, it is known that 71.6 per cent of the wives of the indigent were employed,48 apparently because most women of the lower class had to have extra income to make ends meet. Only if their low wages were added to their husbands' could the urban poor manage to eke out an existence.
Finally, we should address the status of the "non-resident family members" (hi-genju jinko), the majority of whom were employed minors between 10 and 19 years of age. By far the majority of girls in this category were employed as maids or shop-helpers (tetsudai); boys worked mainly as factory workers in the metal and machinery industries. This implies that though they were still under severe constraints, men were on their way towards escaping from their class conditions by working for a big company in heavy industry.49
Table 3.14. Occupations of Female Job-holders, 1912

2. Metamorphosis of the Urban Lower Class: The Post-First World War Period

Government studies on the conditions of the poor, begun in 1911-1912, were launched again a decade later, after the First World War.50 Again, it was popular unrest, culminating in an outbreak of riots known as the rice riots of 1918, which spurred public authorities to action in 1920-1921. But though the events that triggered the poverty surveys of 1910-1911 and 1920-1921 were similar, their findings diverged quite substantially.
In Tokyo's case, the "indigent" had accounted for 12.6 per cent of the population in 1911-1912, but by the 1920 survey they had dwindled to a mere 3.4 per cent.51 It was. not because the criteria for indigence had become any more severe: indeed, they were the same, based mainly on income, during both periods.52 It was simply that far fewer people were under the poverty line.
From about 1917 (midway through the First World War) until March 1920, when the first post-war panic halted the wartime economic boom, the Japanese economy prospered and industrial production soared. Enterprises fought to get the workers they needed to increase production. There was a generalized labour shortage, to the advantage of workers, whose numbers grew remarkably, especially in the metal and machinery industries, where they grew 1.7-fold between 1914 and 1919.53 It was mentioned above that many poor male non-resident family members were employed in the metal and machinery industries in 1911-1912; throughout the First World War boom, a certain number of the urban poor, notably these poor male nonresident family members, managed to rise from the lower class. The result was both quantitative shrinkage and major qualitative transformation of urban poor society. The changes are described below.
First, poverty became geographically decentralized. In Tokyo, its centre of gravity shifted from the traditional poor quarters of late feudal times―Asakusa and Shitaya―to Fukagawa and Honjo. And despite the prevalent indigence of these four wards, their share of Tokyo's total indigent population declined. Thus the number of poor shrank at the same time that poor people began to settle in different districts of Tokyo. "There was gradual improvement in the poor quarters in the centre of the city, and the poor began to move from the centre to the outskirts, then from the outskirts to the suburban and rural districts outside the city."54
Second, workers in big factories, particularly in heavy industry, rose from the ranks of the poor. Eventually a clear line was drawn between workers in heavy industry and the lower strata of urban society according to housing, income level, and participation of family members in several other occupations.55 Those factory workers who were still seen in the ranks of the urban lower class were henceforth employed by small firms.
Third, the urban lower class stopped passing on its poverty to the younger generation. During this period, only 8 per cent of the indigent had had indigent parents.56 Industrial workers, especially workers in big factories, born in urban poverty of many generations' standing, were breaking away from their class origins, with the result that "the younger generation was no longer locked into the self-perpetuating cycle of poverty breeding poverty in the cities."57
Fourth, housing improved. Cheap lodging-houses, where the bulk of urban poor families had lived until the first decade of the 1900s, became primarily the temporary lodgings of unmarried young men, and ceased to house families. Families moved to row houses, which evolved from their original form of communal tenements to apartment houses with a modicum of privacy.
Table 3.15. Occupations of the Indigent, 1920-1921 (percentages)
In the period during which these changes took place (1920-1921), the employment structure of the urban lower class also changed (table 3.15). First, the percentage of male heads of household employed in industrial occupations had declined compared to 1911-1912, while that of manual labourers (coolies, rickshaw-drawers, porters, etc.) had increased. The same shift from industrial to manual labour applied to their wives. Though the majority of wives with jobs (about 60 per cent) were still employed in industry, the percentage engaged in transportation and "other industries" employing manual labour rose in comparison with the previous study period, 1911-1912.
In discussing wives' occupations, it must be borne in mind that the percentage of employed wives dropped substantially after the First World War, though this factor is not indicated in table 3.15. In 1911-1912, 70 per cent of wives of indigent men held outside jobs; in 1921, the rate was 44 per cent.58 To be sure, a greater percentage of urban lower-class wives held jobs than did the wives of workers employed by major heavy industrial companies. But the difference was nevertheless clear with respect to the period when wives had to have their own income to make ends meet. The drop in married women's employment rate is presumed to result from a rise in the income of heads of households. As a result of the wartime labour shortage and a strong labour movement from the middle of the war period on, factory workers' wages rose in both absolute and relative terms from 1918-1919. The incomes of the poor "followed this upward trend" in factory workers' wages, but with a certain time-lag; around 1920-1921 the poor "sought to benefit from wartime changes."59 Under these circumstances, employed wives' incomes rose with respect to 1911-1912 levels, and the average number of days worked sank to 20 a month.
The percentage of employed detached family members besides wives hardly changed, and, as in 1911-1912, the major form of employment was industrial. More women were employed in the manufacture of apparel and accessories, done mainly at home, than in any other type of work. The number of male factory workers was constant. Non-spouse employed dependents, probably mostly children of the head of household and his spouse, were therefore presumably able to escape from the category of urban lower class.
The last observation concerns family size: in 1921, the average size of an indigent household was 4.1 people,60 whereas the 1911-1912 figure was 3.5. The reason for this increase appears to be that more children ceased to leave the family to work. Before, it had been too difficult to support children; after the First World War, however, it was easier for the family to remain together.

IV. Conclusion

Following the period we have just considered, a number of changes, addressed briefly below, affected the situation of the urban poor. The first affects the number of poor. Until the mid-1920s, the lower class continued to represent about the same percentage of the population as in the immediate post-First World War period, but then its numbers took an upswing. This recrudescence of poverty was caused by the financial panic of 1927 and the depression (known as the "Shows Panic") that followed in late 1929. Bankruptcies and a worsening economic climate affected workers at big factories, proprietors of small independent businesses, and tenant farmers. In 1932-1933, the percentage of the population in the lower class again rose above 10 per cent.61 Surveys conducted during the Showa Panic period defined the lower class mainly in terms of their lack of money for paying living expenses, whereas in previous surveys by government authorities two sorts of criteria had defined poverty: the ability to pay living expenses and the subject's residential environment, namely in substandard group housing in the slums. The latter criterion was abandoned in an apparent response to the geographical spread of such people. Indeed, a survey conducted in Tokyo during the Showa crisis indicates that the main pockets of poverty remained in Fukagawa and Honjo wards, but the poor had come to live everywhere else in the city as well.62 In their new settlements, the lower class lived in row houses and one- to four-dwelling apartments that were somewhat more spacious than in the past.
The occupations of the urban lower class during the Showa crisis were nearly evenly divided among four categories: industry, manual labour, miscellaneous employment, and unemployment.63 This signified a decrease in industrial and manual labour, and a corresponding falling back on odd jobs and "no job at all." In other words, those who lost their jobs in industry and manual labour did not necessarily become totally unemployed; a very large proportion of them performed odd jobs, whose availability reduced the direct impact of lost jobs on unemployment figures. However, even odd-job holders were far from able to pay their living expenses, and the majority were regarded by the government as needing assistance. The greatest number of female heads of households were employed in industry, the next greatest in sales. The breakdown of industrial occupations shows that the vast majority were active in "various industries" such as sewing/tailoring and the manufacture of footwear,64 and this probably applied equally to the wives of the poor. Women's occupations remained the same as before. During this period, wives and children again strongly desired to find jobs of whatever kind, even at low wages, because the family budget could not be balanced otherwise.
In sum, during the ensuing economic crisis, a serious reversal took place, again swelling the ranks of the poor and aggravating their living conditions, which had undergone a measure of improvement in the wake of the period of rapid economic growth stimulated by the First World War. The situation of urban lower-class working women during the war economy put in place after the Manchurian Incident (1931)―the next phase of the history of the Japanese urban poor―deserves to be addressed elsewhere.65

Notes

1. Tsuda Masumi, "Nihon no toshi kaso shakai," [Japan's Urban Lower Class], Keizaigaku ronshu, vol. 24, no. 2 (1956); Tsuda Masumi, Nihon no toshi kaso shakai, [Japan's Urban Lower Class] (Mineruba Shobo, Tokyo, 1972).
2. Sumiya Mikio, Nihon chinrodo shiron [Historical Essay on Wage Labour in Japan] (Tokyo University Press, Tokyo, 1955); Sumiya Mikio, Nihon no rodo mondai [Japanese Labour Problems] (Tokyo Daigaku Shuppankai, Tokyo, 1967).
3. Hyodo Tsutomu, Nihon ni okeru roshi kankei no tenkai [The Evolution of Capital and Labour in Japan] (Tokyo Daigaku Shuppankai, Tokyo, 1971).
4. Nakagawa Kiyoshi, "Senzen ni okeru toshi kaso no tenkai" [The Evolution of the Urban Lower Class in Pre-war Japan], parts 1 and 2, Mita gakkai zasshi, vol. 71, nos. 3, 4 (1978); Nakagawa Kiyoshi, "Zoku senzen ni okeru toshi kaso no tenkai" [Sequel to the Evolution of the Urban Lower Class in Pre-war Japan], Shogaku ronshu (Niigata University), 13 (1980); Nakagawa Kiyoshi, "Kanto daijishin chokugo no toshi kaso" [The Urban Lower Class in the Aftermath of the Great Kanto Earthquake], Shogaku ronshu, 15 (1982).
5. "Urban lower class" in this chapter refers to the group in a restricted sense. Readers should refer to Hashimoto Norichika's Chiho toshi no kaso minshu to minshu bodo [The Lower Class of Regional Urban Centres and Popular Uprisings] (United Nations University, Tokyo, 1980) for materials on the lower class in outlying urban areas.
6. Included in Tokyo-shi shiko, 48 (1959). My citation, however, is from Ishizuka Hiromichi, Tokyo no shakai keizaishi [The Social Economic History of Tokyo] (Kinokuniya Shoten, Tokyo, 1977).
7. Yokoyama Gennosuke, Nihon no kaso shakai [Japan's Lower Classes] (Iwanami Bunko, Tokyo, 1949), p. 22.
8. Ishizuka, Tokyo no shakai keizaishi, pp. 23-24.
9. Matsumoto Shiro, "Bakumatsu, ishin-ki ni okeru toshi no kozo" [Urban Infrastructure during the Late Feudal and Early Meiji Periods], Mitsui Bunko Ronso, 4 (1970).
10. Miyachi Masato, Nichi-Ro sengo seiji-shi no kenkyu [The Political History of the Post-Russo-Japanese War Period] (Tokyo Daigaku Shuppankai, Tokyo, 1973), pp.193-200.
11. Yoshida Kyuichi, "Edo jidai no toshi kaso shakai" [The Urban Underclass in the Edo Period], Nihon Shakai Jigyo Daigaku kenkyu kiyo, vol. 27 (1981), p. 29.
12. Ibid.
13. Sumiya, Nihon chinrodo shiron, pp. 83-89.
14. Ibid.
15. Unno Fukuju, "Genchikuron" [Theory of Primitive Accumulation], in Ishii Kanji, Unno Fukuju, and Nakamura Masanori, eds., Kindai Nihon keizaishi wo manabu [Learning from Modern Japanese Economic History], vol. 1 (Yuhikaku, Tokyo, 1977), p. 17.
16. Daiga Koji, Hintenchi kikankutsu tankenki [Exploring the Cold and Starving Slums], 1893; reprinted in Nishida Taketoshi, ed., Meiji zenki no toshi kaso shakai [The Urban Lower Class in the Early Meiji Period] (Koseikan, Tokyo, 1970), 108.
17. Sumiya, Nihon chinrodo shiron, p. 100.
18. Ishizuka Hiromichi, Toshi kaso shakai to saimin jukyoron [The Urban Lower Class and the Living Status of the Indigent] (United Nations University, Tokyo, 1979).
19. Rodo Undo Shiryo Iinkai [Labour Movement Reference Committee], ed., Nihon rodo undo shiryo [Historical References to the Japanese Labour Movement], vol. 1; reprinted in 1962.
20. Kokumin shimbun, 29 October and 5 November 1893; reprinted in Nihon rodo undo shiryo, vol. 1.
21. Reprinted in Nishida, Meiji zenki no toshi kaso shakai.
22. At the 12 December 1981 research meeting of the United Nations University project team, Dr Helena Sumiko Hirata of Brazil pointed out that the collection of trash had become a means of earning income from early in the modern period. Although it would be extremely interesting to study the extent to which trash recyling technology developed in Japan, this question will not be taken up in this study.
23. Tokyo no hinmin serial in Jiji shinpo, 1896; reprinted in Hayashi Hideo, ed., Ryumin [Migrants], vol. 4 (Shinjinbutsu Oraisha, 1971).
24. Oishi Kaichiro, "Kadai to hoho" [Tasks and Methods], in Oishi Kaichiro, ed., Nihon sangyo kakumei no kenkyu [Studies on the Japanese Industrial Revolution], vol. 1 (Tokyo Daigaku Shuppankai, Tokyo, 1975).
25. Ishizuka Hiromichi, Nihon shihon shugi seiritsushi kenkyu [Studies on the History of the Establishment of Japanese Capitalism] (Yoshikawa Kobunkan, Tokyo, 1973), pp. 315-330.
26. Sumiya Mikio, Nihon no rodo mondai [Japan's Labour Problem] (Tokyo Daigaku Shuppankai, Tokyo, 1967), pp. 66-72.
27. At this point in the chapter, the manner of presentation changes, because data are not available on the overall conditions of the urban lower class in specific regions or in the country as a whole. It is only extrapolation of the information in part I and section 1 of part III that allows us to reconstruct the total numbers and the geographical location, occupational composition, and family conditions of the urban poor. Studies of the match industry do give us, however, a specific but partial picture of conditions in urban lower-class society. Originally, a study was also made of the spinning industry, but space does not permit its inclusion here. The most comprehensive treatment thus far of female labour in the spinning industry during the period under consideration is that of Murakami Hatsu, "Sangyo kakumeiki no joshi rodo" [Female Labour in the Industrial Revolution], in Joseishi Sogo Kenkyukai [History of Women Research Group], ed., Nihon joseishi [The History of Japanese Women], vol. 4: Gendai [The Modern Era] (Tokyo Daigaku Shuppankai, Tokyo, 1982).
28. Takamura Naosuke, "Sangyo-boeki kozo" [The Industry and Trade Structure], in Oishi, Nihon sangyo kakumei no kenkyu.
29. Komiyayama Takuji, Nihon chosho kogyo no kenkyu [Studies on Japanese Small Industrial Enterprises] (Chuo Koronsha, Tokyo, 1941), p. 142.
30. Yokoyama, Nihon no kaso shakai, p. 129; Agriculture and Commerce Ministry, Zenkoku kojo tokeihyo [Nationwide Factory Statistics Table] (annual).
31. Yamashita Naoto, "Keiseiki Nihon shihon shugi ni okeru matchi kogyo to Mitsui Bussan" [The Match Industry and Mitsui Bussan during the Formative Period of Capitalism in Japan], pp. 100-105; also the source for the following material.
32. Ministry of Agriculture and Commerce, Commerce and Industry Bureau, ed., Shokko jijo [Conditions of Workers] (1903; reprinted in 1967), p. 129; hereafter this report is referred to as Matchi shokko jijo [Conditions of Match Factory Workers].
33. Ibid.
34. Ibid.
35. Rodo Undo Shiryo Iinkai [Labour Movement Reference Committee], Nihon rodo undo shiryo [Historical References to Japanese Labour], vol. 10 (Rodo
Undo Shiryo Kanko Iinkai, Tokyo, 1959), p. 270.
36. Matchi shokko jijo, p. 167.
37. Ibid., p. 136.
38. Ibid., pp. 136-138.
39. Ibid.
40. Ibid., pp. 134-139.
41. Ibid., p. 151.
42. Miyachi Masato, "Nichi-Ro zengo no shakai to minshu" [Society and the People before and after the Russo-Japanese War], in Rekishigaku Kenkyukai and Nihonshi Kenkyukai, eds., Koza Nihonshi [Lectures on Japanese History], vol. 6: Nihon teikoku shugi no keisei [The Emergence of Japanese Imperialism] (Tokyo Daigaku Shuppankai, Tokyo, 1970).
43. Ishizuka, Tokyo no shakai keizaishi, pp. 101-102.
44. Nihon Shakai Jigyo Daigaku Kyuhin Seido Kenkyukai [Japan College of Social Work Study Group on the Poverty Relief System], ed. Nihon no kyuhin seido [Japan's Poverty Relief System] (Keiso Shobo, Tokyo, 1960), p. 153.
45. Ministry of Home Affairs, Regional Affairs Bureau, Saimin chosa tokeihyo [Statistical Tables from a Study of the Poor], pp. 21-24; Saimin chosa tokeihyo tekiyo [Synopsis of Statistical Tables from a Study of the Poor], pp. 16-17.
46. Nakagawa, "Senzen ni okeru toshi kaso no tenkai," part 1.
47. Ibid.
48. Ministry of Home Affairs, Regional Affairs Bureau, Saimin chosa tokeihyo tekiyo, pp. 16-17.
49. Hyodo, Nihon ni okeru roshi kankei no tenkai, pp. 315-20.
50. In Tokyo's case, such studies included Tokyo-shi Shakaikyoku [Tokyo Metropolitan Social Affairs Bureau], Tokyo shinai no saimin ni kansuru chosa [A Survey of the Poor in Tokyo] (survey conducted in 1920, report published in 1921); and Ministry of Home Affairs, Shakaikyoku, Saimin chosa tokeihyo [Statistical Tables from a Study of the Poor] (survey conducted in 1921, report published in 1922).
51. Tokyo-shi Shakaikyoku, Tokyo shinai no saimin ni kansuru chosa, pp. 4-5.
52. Nakagawa, "Senzen ni okeru toshi kaso no tenkai," part 2, pp. 95-96.
53. Ninomura Kazuo, "Rodosha kaikyu no jotai to rodo undo," Iwanami koza Nihonshi, vol. 18 (Iwanami Shoten, Tokyo, 1975), p. 105.
54. Tokyo-shi Shakaikyoku, Tokyo shinai no saimin ni kansuru chosa, p. 23.
55. Hyodo, Nihon ni okeru roshi kankei no tenkai, pp. 442-79. Big heavy industrial plants were then at a turning point, and beginning to hire part-time workers and have work done by outside subcontractors. What drew the line between workers in these factories and the urban lower class was the permanence of workers' status in big factories.
56. Tokyo-shi Shakaikyoku, Tokyo shinai no saimin ni kansuru chosa, pp. 93-94.
57. Nakagawa, "Senzen ni okeru toshi kaso no tenkai," part 2, p. 99.
58. Ibid., p. 108.
59. Ibid.
60. Tokyo-shi Shakaikyoku, Tokyo shinai no saimin ni kansuru chosa, pp. 23-24.
61. Nakagawa, "Zoku senzen ni okeru toshi kaso no tenkai," p. 29.
62. Tokyo-shi Shakaikyoku, Tokyo shinai yokyugosha ni kansuru chosa [A Survey of People Requiring Relief in Tokyo] (survey conducted in November 1931, published in February 1932), fig. 1, etc.
63. Nakagawa, "Zoku senzen ni okeru toshi kaso no tenkai," p. 31.
64. Tokyo-shi Shakaikyoku, Tokyo shinai yokyugosha ni kansuru chosa, pp. 42-65.
65. This chapter limits consideration to the ethnic Japanese urban lower class, but, beginning in the middle of the First World War, the number of Koreans and other colonized nationalities in the lower strata of Japanese urban society also increased rapidly. To date, the most detailed and systematic treatment of the problems of this segment of the urban poor is found in two works, Matsumura Takao's "Nihon teikoku shugi ka ni okeru shokuminchi rodosha" [Labourers from the Colonies under Japanese Imperialism], in Keio University's Keizaigaku nenpo, vol. 10 (1966) and Totsuka Hideo's "Nihon ni okeru gaikokujin rodosha mondai ni tsuite" [The Foreign Worker Problem in Japan], in the University of Tokyo's Shakai kagaku kenkyu, vol. 25, no. 5 (1974).