The Stagnation of Old Age Pauperism Rates, 1891–1908

A comparison of tables 3b and 4 shows that from January 1, 1892 to March 31, 1906, old age pauperism rates declined only in the four southern divisions. In the midlands, northern divisions, and London, pauperism rates remained constant or increased slightly. For England and Wales as a whole, the old age pauperism rate was nearly identical on the two dates. Why didn't old age pauperism rates continue to decline after 1892? Booth (Reference Booth1894: 514) concluded that the major reason for declining pauperism among the aged during the 1880s was improvements in manual workers’ living standards. The rate of increase in the “prosperity of the working classes” was substantially slower after 1891 than in the previous two decades. From 1891–92 to 1906–7, real full-time wages of manual workers increased by 13.6 percent, after increasing by 37.9 percent from 1871–72 to 1891–92. The increase in wages was even smaller in key sectors—real wages in the building trades increased by only 8.1 percent from 1891–92 to 1906–7, and real wages in agriculture increased by 7.8 percent (Feinstein Reference Feinstein, Scholliers and Zamagni1995: 260–61, 264–65).Footnote ¹⁰

The pressure on living standards was most severe among low-skilled workers. The best available indicator of the effect of the trade cycle on the very poor is the rate of indoor pauperism among adult able-bodied males. Able-bodied workhouse inmates were from the “bottom of the social hierarchy.” While the number of male able-bodied workhouse inmates was small, it was very “responsive to indicators of general economic conditions” (MacKinnon Reference MacKinnon1986: 330–34). MacKinnon calculated the trend in male able-bodied indoor pauperism for three broadly defined regions—the North and Midlands, the South, and London.Footnote ¹¹ Pauperism increased in the early 1890s in each region, then declined somewhat in the late 1890s before increasing again after 1900. For the entire period 1887 to 1908, the share of able-bodied males relieved in workhouses increased by 131 percent in the North and Midlands, 63 percent in London, and 48 percent in the South. The data suggest that employment opportunities for the poorest 10 to 20 percent of the population (those most likely to turn to the Poor Law for assistance in old age) deteriorated during the two decades leading up to World War I. The extent of deterioration was largest in the North and Midlands, although job opportunities for the low-skilled declined throughout England. This, combined with the meager growth in wage rates, provides an explanation for the trend in old age pauperism from 1891 to 1908.

As noted in the preceding text, much of the decline in old age pauperism after 1871 was a result of a change in policy rather than improving economic conditions. MacKinnon estimated that if the Crusade Against Outrelief had not occurred as many as “200,000 more old people would have been on outdoor relief in 1900” (1984: 327–28). While this is almost certainly an upper bound estimate of the impact of the crusade, adding 100,000 to the number of relief recipients 65 and older in 1908 yields an old age pauperism rate of 31.7 percent, only six percentage points lower than the pauperism rate in 1871. The economic growth in late Victorian England did not lead to a sharp reduction in poverty among the aged.

The estimates in tables 2–8 indicate that the Poor Law continued to play a major role in the lives of working-class individuals in late Victorian England despite its increasing stinginess and the increase in workers’ real wages over time. The data clearly show that large-scale government assistance for the elderly is not a product of the welfare state; it existed at the local level throughout the nineteenth century and earlier.

Gender Differences in the Number and Treatment of Paupers

There were substantial differences in male and female pauperism rates and in the methods used to relieve male and female paupers in Victorian England. The day-count and 12-month estimates of pauperism rates for men and women aged 65 and older in each registration division in 1871, 1881, 1891, and 1908 are presented in table 9. The one-day pauperism rates were higher for women than men in all registration divisions except London throughout the period. The gender differences were largest in the more rural regions—the East, South West, and North. The relationship between the 12-month male and female pauperism rates is less clear cut. Column 1 shows that the ratio of the number of paupers who received relief at some point over the 12-month period ending on March 25, 1892 to the day count for January 1, 1892 was larger for old men than for old women in all ten registration divisions. The ratio was largest in London and the industrial North West, and lowest in the rural South West, South Midlands, and East. This suggests that aged women were more likely to be “permanent” paupers who received relief week in and week out, while aged men were more likely to receive relief for only part of each year.Footnote ¹² As a result, the differences in male and female 12-month pauperism rates are smaller than the day-count differences, and after 1881 male pauperism rates were higher than female rates in London and the industrial North West.

TABLE 9.

Pauperism rates for men and women 65 and older, day and year counts

Sources: Data on the annual number of non-able-bodied male and female paupers for each registration division from the annual Reports of the Poor Law Board (for 1861) and the Local Government Board (for 1871–1908).

Note: The ratio of the year count to the day count of aged paupers for London in 1891–92 is far larger than that for any other division, and almost certainly overestimates the extent of London pauperism. In column 1, I set the ratio for London equal to that for the North West.

Table 9 also shows the changes in male and female pauperism rates over time. From 1871 to 1908, the female pauperism rate (both one-day and 12-month) declined by more than the male rate for all ten registration divisions.Footnote ¹³ For example, in the South Midlands the male one-day pauperism rate declined by 10.5 percentage points from 1871 to 1908 while the female one-day rate declined by 19.6 percentage points; in the North West the male pauperism rate declined by 2.2 percentage points and the female rate declined by 11.1 percentage points. The decline in both male and female pauperism was especially pronounced during the Crusade Against Outrelief in the 1870s.

The form of relief offered to the aged differed substantially between men and women. Table 10 shows that far more old men than women were relieved in workhouses throughout the late nineteenth century. In 1871 more than 20 percent of male paupers 65 and older were relieved in workhouses in eight registration divisions, and more than 40 percent of male paupers were relieved indoors in London and the North West. By contrast, eight registration divisions relieved fewer than 10 percent of aged female paupers in workhouses, while London and the Northwest relieved more than 15 percent of their aged female paupers indoors. The male-female differential became even more pronounced following the Crusade Against Outrelief. Excluding London, the share of men 65 and older relieved in workhouses increased by 12.1 percentage points from 1871 to 1881, while the share of old women relieved in workhouses increased by 4.4 percentage points.Footnote ¹⁴ According to the pauper census of January 1, 1892, six registration divisions relieved more than one-third of their aged male paupers in workhouses, while only three divisions relieved more than 15 percent of aged female paupers indoors. Despite changes over time in the use of the workhouse, those registration divisions that relieved a relatively large (small) share of males in workhouses also relieved a relatively large (small) share of females in workhouses throughout the period from 1861 to 1908.

TABLE 10.

Percentage of male and female paupers 65 and older relieved in workhouses

Sources: Data on the percentage of non-able-bodied male and female paupers relieved in workhouses for each registration division from the annual Reports of the Poor Law Board (for 1861) and the Local Government Board (for 1871–1908).

There were several economic reasons for the male-female differential in pauperism rates. Old women were far less likely to be employed than men and were paid lower wages. A “test census” of persons aged 65 and older in 28 representative Poor Law unions conducted by the Departmental Committee on the Aged Deserving Poor found that nearly three-quarters of women 65 and older had weekly incomes from all sources of 10 shillings or less, as compared to 46 percent of men.Footnote ¹⁵ Booth's 1892–93 survey of rural districts found that nearly two-thirds of men aged 65–70 and 46 percent of men aged 70–75 had some wage earnings. By comparison, only 41 percent of women aged 65–70 and 19 percent of women aged 70–75 reported any wage earnings. Old women also were less likely to have any savings or benefits from friendly societies (Booth Reference Booth1894: 346–49).

Old women were far more likely to obtain assistance from relatives or friends than old men. The 1899 test census found that 70 percent of women 65 and older with income less than 10 shillings and not in receipt of poor relief received help from children, other relatives, or friends, as compared to 28 percent of old men. These large differentials occurred in urban, rural, and mining districts. Booth's survey found that 27 percent of women and 18 percent of men 65 and older received help from relatives or charity. The male-female differential declined with age—the share of men and women aged 80 and older who received assistance from relatives was very similar: 32 percent and 35 percent.

Why were aged male paupers more likely to be relieved in workhouses than aged female paupers? Local officials denied that they treated old men differently than old women. Some unions claimed to make indoor/outdoor relief decisions based on the character of the applicant, and there is evidence that women were more likely than men to be judged of “good character.” In his testimony before the Royal Commission on the Aged Poor, Booth stated that one reason why so many more old women than men were granted outdoor relief was because “a very large number of [women] are decent respectable people.”Footnote ¹⁶

Old people who lived with relatives were able to get by on a small amount of outdoor relief, and there is some evidence that children were more willing to have their aged mothers live with them than their fathers. Anderson (Reference Anderson, Shannas and Sussman1977: 51) maintained that “old people who were able to make some contribution to the family with whom they lived were most likely to be welcomed, while those who imposed a particular strain on the family were most likely to be rejected.” Old women typically were better able to perform useful domestic functions, such as child minding, cooking, and cleaning, than were old men. Crowther (Reference Crowther1981: 234) suggested that children often were willing to assist their mothers but not their fathers because “affectionate ties with the mother were likely to be stronger.” Some contemporaries maintained that old men were more willing to enter workhouses than old women, who “would endure much greater hardships before giving up the struggle for independence” (MacKinnon Reference MacKinnon1984: 110–11). In sum, the male-female differential in workhouse relief was the result of a combination of administrative, economic, and social causes.

During the Crusade Against Outrelief in the 1870s, the female pauperism rate declined by more than that for males, while the share of men relieved in workhouses increased far more than that for women. How can these two results be reconciled? Is this evidence that local Poor Law authorities treated old men and women differently? Suppose that the same share of men and women were offered workhouse relief during the crusade. If old men were more likely to accept indoor relief while women were more likely to refuse relief altogether rather than enter a workhouse, we would observe both a greater decline in the share of old women on relief and a greater increase in the share of old men being relieved in workhouses. The same result would be observed if children were willing to assist their mothers, but not their fathers, rather than have them enter a workhouse. It seems reasonable to conclude, along with Thane (Reference Thane1996: 34), that “the change of policy [initiated by the crusade] hit women especially hard since they had greatly outnumbered men as recipients of outdoor relief.” This was one of the (presumably) unintended consequences of the crusade, the sort of effect that economists love to point out.

Regional Differences in Economic Circumstances

Most contemporaries agreed that differences in the availability of employment were a major reason for regional differentials in old age pauperism. Booth (Reference Booth1894) maintained that old workers were better able to find full- or part-time employment in agriculture than in urban areas, based on the responses he received to a questionnaire on the condition of the local aged population sent to clergy throughout England and Wales. Many of the responses for urban districts noted the absence of work for the aged. In Dewsbury, “men after 60 find work very scarce; mills do not employ them unless skilled,” while in Birmingham “the chances of the aged getting casual labour are restricted by the competition of those between 45 and 60.”Footnote ²¹

The responses from rural unions were more mixed, but many reported that work for healthy old people was plentiful. Several unions reported that “field work is available for men as long as they are able to do it.” In many agricultural districts jobs existed for the old because of the outmigration of younger workers—one union reported that “labourers are scarce, and the industrious, though aged, are not lightly parted with” (ibid.: 150–77, 262–309). Booth concluded that in rural districts “old men keep their vigour longer, and even when feeble or ailing can still obtain and do some agricultural work.” By contrast, in towns old men “break down at an earlier age,” and are the victims of “increased competition for such work as . . . [they] are able to perform.” In sum, the “effective working life is ten years longer in the country than in the town, or, speaking generally, is as seventy to sixty” (ibid.: 321–22).

J. A. Spender (Reference Spender1892: 23–46) found that for many urban trades a worker's wage declined sharply after age 55, and even earlier for dock workers, building trades’ laborers, and other low-skilled workers employed in heavy labor. Wages of agricultural workers, while lower than those of urban low-skilled workers, often did not decline until workers reached the age of 65. However, there were fewer opportunities for casual employment in rural areas than in towns, and the condition of farm laborers over 65 “is generally one of desperate poverty” (ibid.: 41–43).

Historians disagree as to the relative availability of jobs in rural and urban districts. Thane (Reference Thane2000: 275) contends that “old men could work longest in agriculture,” and Macnicol (Reference Macnicol1998: 49–50) concludes that agricultural districts provided older workers with “a myriad of casual or light employment opportunities” that enabled them to continue working past age 65. Conversely, Hunt (Reference Hunt1989: 418) maintains that employment for the aged generally was more plentiful in urban than in rural districts, so that “there were more opportunities for old men and women to earn at least part of their own maintenance” in towns. Johnson (Reference Johnson1994: 114, 120) argues that the occupational data reported in the decennial censuses of 1881 to 1911 do not support the notion “that it was becoming increasingly difficult for [urban] workers to obtain a position after the age of 50.” He also found evidence, however, that many urban workers in old age switched from jobs in high-paying sectors to jobs in low-paying sectors.Footnote ²²

Some idea of the age at which employment rates for old people began to decline can be obtained by examining pauperism rates by age. Table 11 reports pauperism rates for men aged 60–65, 65–70, and 70 and older on March 31, 1906 for each registration division. The pauperism rate for men 70 and older was higher in London and the other four southern divisions than in the three northern divisions. This was not the case, however, for men under age 70. The pauperism rate for men aged 60–65 and 65–70 was lower in the South Midlands and South West, both relatively rural divisions, than in the North West, Yorkshire, or North. The rural-urban differences were even more pronounced at the county level. The pauperism rate for men aged 60 to 70 was 6.2 percent in the rural agricultural county of Suffolk, as compared to 8.3 percent in the urban/industrial county of Lancashire. For men in their 70s or older, the relationship was reversed—the pauperism rate was 27.6 percent in Suffolk and 16.0 percent in Lancashire.

TABLE 11.

Male pauperism by age, March 31, 1906

Source: Parliamentary Papers, Royal Commission on the Poor Laws and Relief of Distress, Appendix Vol. XXV, Statistical Memoranda (1910, LIII), 176–93.

The data suggest that job opportunities for men in their 60s were relatively good in agriculture, but that employment declined sharply after age 70. The fact that pauperism rates for men in their 60s were higher in industrial/urban districts offers some support for the notion that urban workers lost their jobs at a younger age than rural workers. However, the relatively low pauperism rates for men in their 70s suggest that urban areas offered more opportunities for the aged to find casual employment. The exceptionally high pauperism rate for London men supports the widely held idea that the London labor market suffered from chronic oversupply.

Booth (Reference Booth1894: 16–19) grouped Poor Law unions into six industrial categories, and calculated pauperism rates for persons 65 and older in 1891–92 for a sample of unions in each category.Footnote ²³ For England as a whole, old age pauperism was lowest in residential unions, highest in seaports, and similar across manufacturing, mining, and agricultural unions. For each industrial category, pauperism rates were lower in the North than in the Midlands, East, or South. The north-south differential was especially large in purely agricultural districts, where the old age pauperism rate in the North (15.5 percent) was less than half that in the East (34 percent) or South (32.5 percent). The regional differences in pauperism rates within industrial categories were larger than the differences across industrial categories.

The differences within industrial categories were caused to some extent by regional wage differentials. Wages in agriculture and the building trades generally are considered to be representative of rural and urban wage rates (Bowley Reference Bowley1900: 59, 63; Hunt Reference Hunt1973: 4–5). Weekly wage rates for agricultural laborers and bricklayers’ laborers varied substantially across registration divisions. In agriculture, wages were highest in the three northern divisions and lowest in the South Midlands, East, and South West.Footnote ²⁴ Nominal wage rates in the building trades were highest in London, followed by the northern and midlands divisions, and were lowest in the South Midlands, East, and South West. The higher wage rates in northern districts enabled workers to save more and to join friendly societies, and enabled those who continued to work after age 65 to be better able to get by without applying for poor relief. In addition, higher wages enabled children of the aged who lived near their parents to offer them more financial assistance.

Regional differences in economic circumstances effected pauperism rates in one other way. Children were more likely to assist their aged parents if they lived near them. Other things equal, old people in districts with low outmigration rates would be more likely to have children living nearby, and therefore more likely to receive help from their children, than those in districts with high outmigration rates. Outmigration rates were especially high in the low-wage, and rural, South Midlands, East, and Southwest.Footnote ²⁵ As a result, a larger share of old people in the south than in the high-wage and more urban north had no children to whom they could turn for assistance, and were forced to apply for poor relief when they could no longer support themselves.

Booth (Reference Booth1894: 339–52) analyzed the sources of maintenance for the elderly in rural districts in 1892–93, based on a survey of 9,125 persons aged 65 and over (4,450 males and 4,675 females) living in 262 rural parishes located throughout the country. For England and Wales as a whole, nearly half of the elderly had some wage earnings, although Booth stressed that these earnings often were quite meager. Slightly more than a third had savings, and 55 percent maintained themselves entirely through earnings and savings. Some 22 percent of the elderly received outdoor poor relief, although only 5 percent relied entirely on the Poor Law. The importance of the various sources of maintenance differed across regions. The two extremes were, not surprisingly, the high-wage north and the low-wage east. In the north, 45.8 percent of the elderly had some savings, 63.3 percent maintained themselves entirely through earnings and savings, and only 13.4 percent received outdoor relief. By contrast, in the east only 26.1 percent of the elderly had savings, 49.4 percent maintained themselves through earnings and savings, and 28.6 percent received outdoor relief.Footnote ²⁶

In sum, most aged individuals had income from one or more sources, including wages, savings, friendly society or trade union benefits, or help from children, other relatives, or friends. For many, however, income from these sources was less than they needed to subsist, and they were forced to turn to the Poor Law for help.

Regional Differences in Relief Administration

Regional differentials in pauperism rates were not the result of economic forces alone—the local administration of poor relief also played an important role. The most important difference across unions in relief administration was the use of the workhouse. Some unions offered workhouse relief to virtually all aged paupers, while others gave outdoor relief to those aged persons who were able to maintain themselves at home. Between these two extremes there was a continuum of policies. In many unions officials attempted to determine the character of relief applicants, and offered outdoor relief to those deemed to be “deserving” and workhouse relief to those deemed “undeserving.”

All agreed that the aged attached a stigma to indoor relief and that, other things equal, the offer of the workhouse lowered pauperism rates. One person who testified before the Royal Commission on the Aged Poor commented that “there is the strongest objection on the part of those who have a home at all, or can find a home among any of their relatives, to enter the workhouse.” The poor's objections to the workhouse were not necessarily those portrayed by early and mid-Victorian writers such as Charles Dickens—by the 1860s workhouses had “ceased to threaten cold, hunger, and mistreatment” (Crowther Reference Crowther1981: 246). Sir Hugh Owen testified that “the great majority” of aged paupers in workhouses were “better fed, better housed, and better provided for than they ever were before.” The poor objected to the discipline of the workhouse and to their loss of freedom—a witness before the Royal Commission stated that the poor “look upon the workhouse as a kind of prison.” Moreover, the acceptance of workhouse relief threatened “social disgrace” both for the pauper who accepted it and for his or her family.Footnote ²⁷

Can differences in relief administration explain the regional differentials in old age pauperism rates? Outside of London there was a negative relationship between the use of workhouses and the pauperism rate—the share of aged paupers relieved in workhouses was higher in northern than in southern divisions. The deterrent effect of workhouse relief has been empirically estimated for paupers of all ages in 1868–69 by MacKinnon (Reference MacKinnon1987) and for persons aged 65 and older in 1891–92 by Boyer and Schmidle (Reference Boyer and Schmidle2009). MacKinnon found that for nine of ten registration divisions—the North West being the exception—increased use of the workhouse led to a reduction in the number of paupers. The magnitude of the deterrent effect was large—a 10 percent increase in the share relieved indoors led to a 2.5 to 6.0 percent decline in the pauperism rate (MacKinnon Reference MacKinnon1987: 621).

Boyer and Schmidle estimated regression equations to explain the variations in old age pauperism rates across all 585 English Poor Law unions. Their independent variables included the percentage of paupers relieved in workhouses, a series of dummy variables for the industrial character of the union, and a series of regional dummies; for a subset of rural unions, the weekly wage paid to agricultural laborers also was included.Footnote ²⁸ Table 12 reports the estimated coefficients for the most relevant variables from three regressions—one including all 585 unions, one excluding the 29 London unions, and one for a subset of rural and semirural unions. The workhouse had a large deterrent effect in each regression—a 10 percent increase in the share of paupers relieved in workhouses led to a 3.5 to 3.6 percent decline in old age pauperism. The coefficients on the dummy variables for industrial character were small and are not reported in the table. Following Booth, the unions were divided geographically into seven regions. The Industrial North is the excluded region in the regressions. Other things equal, pauperism rates in the East, South, and South West were significantly higher than in the North or Industrial North; pauperism rates in the Midlands were above those in the northern regions but below those in the south.

TABLE 12.

Regressions to explain old age pauperism in 1891–92

Note: Dependent variable is the share of persons aged 65 and older in receipt of poor relief. Standard errors are in parentheses.

Column 3 gives regression results for a subset of rural unions for which agricultural laborers’ wage data are available. The agricultural wage had a strong effect on pauperism—a 10 percent increase in the wage led to a 5.2 percent decline in the old age pauperism rate. Including farm wages reduces the magnitude of the coefficients on the regional dummies somewhat, but they remain significantly different from zero. In other words, the north-south differentials in old age pauperism rates remain even when differences in relief administration, wage rates, and the industrial nature of unions are taken into account. The regional dummies might be serving as proxies for differences in the availability of work outside of agriculture that were not picked up by the “industrial character” dummies. Job opportunities within agriculture also differed across regions—winter employment for the aged must have been especially scarce in the grain-producing South and East (Goose Reference Goose2005: 363–67, 371–76). By contrast, the dummy variables might be picking up noneconomic differences across regions.

Regional Welfare Customs

Some Victorian commentators and historians have offered another explanation for the north-south differentials in pauperism. Booth (Reference Booth1894: 24) concluded that the low pauperism rates in the north were “undoubtedly due to the character of the people,” and other contemporaries agreed. Northerners were deemed to be more likely to have savings accounts and be members of friendly societies than southerners, and to attach a greater stigma to applying for poor relief (Rose Reference Rose and Donajgrodzki1977). More recently, King (Reference King2000: 267–69) has argued that there were “very substantial . . . differences in entitlement, relief generosity, and relief sentiment” between northern and southern England. Not only were southerners more likely than northerners to apply for poor relief, but local relief administrators were more willing to offer them assistance.

What caused these regional differences in welfare customs, and when did they begin? King (ibid.: 268) maintains that the differences date back to the early eighteenth century if not before—the stingy north and west had “a complex religious history and arguably the most distinctive and potentially puritanical cultures,” and also had more developed systems of private charity. Billinge (Reference Billinge, Baker and Billinge2004) suggests that distinct regional cultures began to develop in the second half of the eighteenth century, largely as a result of the uneven geographic nature of the Industrial Revolution.

Distinct regional attitudes toward welfare might have emerged during the heyday of the Old Poor Law from 1795 to 1834. Regional differences in the role of poor relief grew in the late eighteenth century, when the deteriorating economic environment for southern farmworkers led to a sharp increase in the demand for public assistance in the grain-producing south and east. The relief of able-bodied laborers and their families was so widespread during the first third of the nineteenth century that the stigma associated with receiving poor relief must have declined. Southerners became accustomed to being on relief, and many came to believe that they had a right to public assistance during hard times.Footnote ²⁹

The elderly in Victorian England had grown up under the Old Poor Law—many of those raised in the south and east had lived in families that received child allowances from the parish. The customs established before 1834 might have continued to influence the behavior of the aged poor into the 1890s—the shadow cast by Speenhamland was quite long. However, regional differences in attitudes toward welfare should have declined in the last decade of the nineteenth century, as the number of individuals who remembered the period before 1834 fell. This would help to explain the convergence in old age pauperism rates after 1891.Footnote ³⁰