![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
Abstract
A variety of taxes levied at various dates between the 1540s and the early nineteenth century are analysed to ascertain whether the frequency with which widows and unmarried women are listed as taxpayers matches the ratio between the numbers of widows and unmarried women and the number of men in the adult population. Examination of the tax assessments indicates that the proportion of taxpayers who were women often fell well below their proportion in the general population. This applies in the case of all taxes levied between the sixteenth and early nineteenth centuries and regardless of whether a tax was assessed primarily on goods, houses occupied, or land owned or occupied. That double the proportion of female than of male householders were exempted from payment of the hearth tax on houses they occupied in the second half of the seventeenth century indicates that many women maintained households on very depleted resources. However, the proportion of taxpayers who were women amongst those paying the least tax did not usually exceed the proportion of women amongst those paying most tax.
Acknowledgement
This research was supported by a research grant R 000 23 9190 from the Economic and Social Research Council with further support from the Department of History, University of Essex.
Notes
1 This programme was developed by Jim Oeppen of the Max Planck Institute for Demographic Research in Rostock and has been used in this article to estimate the proportion of widows and unmarried women relative to the number of adult men in English populations in the past. These estimates are inferred from the prevailing rates of fertility, nuptiality and mortality and trends in age at first marriage indicated in family reconstitutions of English parish registers.
2 Another factor that could in theory account for lower proportions of widows in the tax-paying than in the general population would be significantly lower rates of widowhood in the tax-paying population. This might arise because women in the tax-paying population experienced higher rates of mortality than poorer women and that therefore fewer of them survived to experience widowhood. Alternatively, it could be argued that husbands who were taxpayers might be more likely to out-live their wives than did poorer husbands as a result either of higher male (but not female) life expectancies in the tax-paying population or earlier male (but not female) marriage in the tax-paying population. However, comparisons of one elite population in England (the Peerage) with the reconstituted populations of English parishes between the seventeenth and early nineteenth centuries, indicate that the Peers were significantly older than their wives at first marriage (usually by 4-5years compared with 1-2years in the reconstituted populations) and that the wives of Peers were more likely than were married women in the general population to spend part of their lives as widows. These calculations by Jim Oeppen are included in CitationMoring & Wall, 2008.
3 Properties held by a husband in the name of his wife are not distinguished from other properties that he owned although exceptionally the land tax assessment for Barton, Westmorland in 1773, lists two assessments of men for land held on behalf of their wives. For a similar explanation for the absence of married women from persons assessed for the subsidies see CitationSchofield, 2004, p. 108.
4 This is the median percentage for the 18 surviving lists and assumes that all women whose marital status was unspecified were never-married. If women of unknown marital status are excluded, the median percentage rises to 90%. Only those lists that identified at least one unmarried woman heading a household have been included. Lists with the requisite information over-represent urban parishes, many with small populations. Comparison is also possible with the micro-simulation which suggests that in the second half of the seventeenth century about one third of women not currently married had never been married. However, the micro-simulation covers a broader section of the population than does the hearth tax in that it would include both servants and adult women living with their parents.
5 Examination of rate-books for particular parishes also reveals some evidence of multiple entries, probably resulting from an existing ratepayer having acquired an additional property.
6 Analysis of land tax returns for a set of adjoining parishes in Sussex in 1785 shows women constituted 12.8% of landowners and that they owned 12.6% of the properties that were separately listed. Women constituted a smaller percentage of occupiers (8.9% and held an even smaller percentage of properties (7.1%). The parishes selected for analysis are listed in Appendix A.
7 Selection was determined by availability of published transcripts in Cambridge University Library and in the Library of the Cambridge Group. Specific sources are detailed in Appendix A. A more extensive search would no doubt extend the coverage.
8 This account is based on CitationFarraday's (1972) introduction to his edition of the Herefordshire militia assessments. Some tenants at rack rent might be listed when the owner was a non-resident (CitationFarraday, 1972, p. 4).
9 Results are presented for administrative divisions of counties (Hundreds) rather than for individual parishes (towns apart) to minimise the impact of random variations due to small populations.
10 Results represent medians for the two groups of districts, distinguished according to the number of taxpayers.
11 An alternative estimate which also excludes all unmarried men and women under the age of 30 as they would be less likely than other adults to hold property suggests that widows constituted 22.1% of the remaining adults in the national population and widows and unmarried women over 30, 23.1%. The latter still far exceeds the percentage of women in the tax-paying population.
12 Although Zeithofer's in his recent study of peasants in Bohemia between 1640 and 1840 reported that the names of widows who had run a farm for several years were not entered in the land registers (CitationZeithofer, 2004, p. 80).
13 There have been suggestions that properties may have been under-valued (see for example CitationFarraday, 1972, pp. 4, 9) but properties held by women would have to be under valued relative to those held by men to undermine the argument that even within the taxpaying population women tended to own less property.
14 In the populations included in , women constituted 6% of property owners paying on average (median) less than 2 shillings (subsidy) or less than £2 (militia) while they constituted 9% on average of those paying over 5 shillings in the subsidy or over £5 for the militia. Properties held by women could of course still have been under-valued (see note 13).
15 Examination of surviving returns, however, shows that some landlords did pay the tax on behalf of their tenants (CitationParkinson, 1994, parish of Rudrisse and for examples of different persons sharing payment of the Hearth Tax see ).
16 This was the case for example in Belper where 11% of taxpayers were women compared with 9% in the rest of the Appletree Hundred of Derbyshire, 15% of taxpayers in Okehampton, Devon, against 10% in the rest of the Hundred of Lifton, and 16% of taxpayers in Merthyr but only 11% in the remainder of the Hundred of Senghenydd. On the other hand in the Hundred of Northwich (Cheshire), the percentage of women assessed to the hearth tax was lower in the town of Sandbach (12%) than in other parts of the Hundred (15%).
17 Averages represent the medians for the 18 surviving lists (see note Footnote4) and the hearth tax populations listed in . In the two areas (Radfield and the Isle of Wight) where the marital status of those women liable to pay the hearth tax was usually specified just under 9% of persons assessed were widows. Comparison of the proportion of women of those assessed to the hearth tax in Lichfield in 1662 with the proportion of households headed by women in the list of 1692 also reveals substantially more households headed by women, and by women noted as widows in the 1692 list than assessed to the hearth tax thirty years earlier (22 and 16% respectively in 1692 compared with the 10 and 7% of those assessed to the hearth tax). The evidence of the lists is in this case is preferred over the micro-simulation for the purposes of comparison as the latter does not identify sub-populations such as household heads.
18 The median for the 13 populations listed in . In Lichfield the differences between the proportion of women householders recorded in the hearth tax returns as either exempt or liable and the proportion of households headed by widows according to the 1692 list remained considerable (16% in the hearth tax return against 22% in the list).
19 For example in Colyton the exempt are all recorded as paupers. However, this is also true of other Devon Hundreds with more modest proportions of exempt persons.
20 The average represents the median of the populations included in .
21 Some of course would never have married or been widowed. For the period 1650-1699 the micro-simulation indicates that the currently married constituted 71% of the male population above the median age at marriage excluding unmarried males under 30 on the grounds that most of these would not have been householders.
22 Details of liability for payment of poor rates are, for example, set out in CitationBott, 1793, vol. I, Section vii, pp. 90-200. Comparison of rate-books and hearth tax returns from 1664 for some parishes in the town of Cambridge revealed that there were more persons listed in the latter than the former even when the exempt are excluded from the comparison, but that 80% of persons rated by the Overseers of the Poor appeared also in the hearth tax returns and over three quarters of those rated by the Churchwardens, see CitationEvans, 2000, p. xlvi.
23 Different rates in the £ might be set for houses and land.
24 For Colyon, se the analysis of the subsidy, above and for the Hundred of Colyton, .
25 The variation is only from 14% to 25% with no clear temporal trend. Instances where the sex of the owner/occupier could not be determined (for example where property was held by trustees, executors, unidentified heirs or institutions), have been excluded. Including the property held by institutions lowers very slightly the proportion of property owned by women.
26 Between 1643 and 1658 there were between 81 and 85 ratepayers depending on the year and between 1663 and 1730, between 97 and 110 again depending on the year.
27 In Westbury, less than 20% of the ratepayers were women in 1666, 1667 and 1670 and between 1680 and 1685; women were more than 30% of the total in 1657, 1668 and 1670. In Stoke Bishop the percentage of women listed in 18 rate-books was less than 20% but exceeded 30% in 1668 and in 1676 and 1677.
28 Photocopies of rate-books were consulted in the Library of the Cambridge Group.
29 Sixteen percent of women ratepayers in 1819 were unmarried, 22% in 1828 and 11% in 1835 (23% of women rated as property owners). The 1726 list did not identify unmarried women.
30 Cambridgeshire Record Office P 129/11/3. The effect of excluding owners and occupiers whose sex could not be determined (see above note Footnote25) on the calculation of the percentage of properties which women owned or occupied is considerably greater (particularly on the ownership of property) in the nineteenth century. For example, analysis of the Over rate-books from 1851 show women owning13% of rated properties when such properties are excluded but only 8% of all properties in the parish.
31 HO 107/1749 Census of Over 1851. Analysed from microfilm in Cambridgeshire Record Office.
32 The segments distinguish four groups of ratepayers from those paying most rates to those paying the least. Unfortunately, it was impossible to define segments which contained equal numbers of ratepayers as the assessments of some large groups of ratepayers were identical. Further examination indicated in four of the populations examined (Luccombe, Stoke Bishop, Shirehampton and Over) there was no sign of a correlation between the proportion of ratepayers assigned to the highest and lowest sectors of the rate-paying population and the percentages of ratepayers who were women. The one exception occurred in Westbury where the percentage of women among the group of ratepayers paying the lowest rates fell steadily in the second half of the seventeenth century while at the same time this sector of the rate-paying population expanded from under a third to more than 40%.
33 In the period as a whole, women were one in every five of those paying the most rates, 27% of the next most highly rated group but only 5 and 16.5% of the two groups of ratepayers in the lowest rated properties. The number of ratepayers in each segment varied from year to year but averaged 15 in the group paying the most rates, 29 in the next most highly rated group, 32 in the third group and 14 in the group paying least rates.
34 Women in Westbury constituted 15% of the most highly rated group and 26.5% of the group of ratepayers with the lowest rates due. In Stoke Bishop, women constituted 18.5% of the most highly rated group and 22% of the lowest rated group, and in Shirehampton 40.5% of those paying most rates and 33.5% of those paying least rates. All percentages are medians of percentages in these segments of the rate-paying population between 1656 and 1697 (1684 in the case of Shirehampton) as aggregated into eight time periods (six for Shirehampton). The number of ratepayers in each segment again varies over time but for the highest rated group, averaged 53 in Westbury, 54 in Stoke Bishop and 35 in Shirehampton. Those paying least averaged 35 in Westbury, 51 in Stoke Bishop and 44 in Shirehampton.
35 Women constituted just 4% of those paying most rates, 8% of the next most highly rated group, 7% of the third group, and 11% of those paying least in rates. The average number of ratepayers in each group between 1757 and 1803 were, respectively, 26, 32, 28 and 24. In 1851, 1881 and 1901 there were fewer than 100 ratepayers in each group on only one occasion (for the highest rated group in 1901).
36 Analyses of rate-books for Colyton from 1726, 1819, 1828 and 1835 (copies in the Library of the Cambridge Group) and of Ardleigh rate-books from 1793 and 1803 (held in the Essex Record Office, D/P 263/11/1).
37 Although it appears there was no legal barrier to their serving in the capacity. For evidence of four women serving as Overseers of the Poor in Ardleigh, during the late eighteenth and early nineteenth centuries, see CitationErith, 1978, p.7. All four were widows of yeomen farmers.
38 See note Footnote6. The ideal of course would be to calculate the full extent of an individual's property holding wherever located but this is precluded by difficulty of determining whether persons of the same name holding property in different parishes are indeed the same person. Within a given parish correct identification is more likely particularly when at least property owners are listed alphabetically (as in the Sussex returns). However, a parish based study can only offer a partial test of the impact of property fragmentation on the proportion of property owned or occupied by women. The implicit assumption made in this study is that if analysis of the fragmentation of property at parish level establishes that it had only a minimal impact on the percentage of properties either owned or occupied by women, the same effect is likely to be present if the property holdings of individuals could be traced beyond parish boundaries However, this assumption has yet to be tested.
39 This has been inferred from the fact that the proportion of properties occupied by women is less than the proportion of occupiers who were women (see above note Footnote6).
40 Copies of these land tax returns (together with many others) were initially collected by the Cambridge Group to provide information on landholding patterns in those parishes and townships selected for demographic research, literacy as measured by the proportion of grooms and brides who signed the marriage registers, and had local censuses taken before 1801 of interest for the information they provided household structure. The assessments used are listed in Appendix 1 and do not constitute a random sample (although they include some of Roger Schofield's national sample of parishes which he used to study literacy). The copies of the assessments are currently housed in the Library of the Department of Geography in the University of Cambridge, and the originals in the respective county record offices.
41 An alternative estimate which excludes all unmarried men and women under the age of 30 as less likely than other adults to hold property, suggests that widows constituted 16.4% of the remaining adults in the national population and widows and unmarried women over the age of 30, 17.4%. The latter still exceeds by a large margin the percentage of women among owners and occupiers of property listed in the land tax records.
42 The Sussex returns have had to be analysed separately as they recorded rental values rather than the tax charged as in the case of the other land tax returns. The rental values as reported are likely to be notional and do not indicate the actual value based on rent paid (see CitationGinter, 1992, p. 113).
43 CitationGinter (1992, p. 663, note 20) counselled against aggregating tax assessments across parish/township boundaries because valuations reported for different parishes are not comparable. This advice has been ignored here because of the small size of some taxpaying populations.
44 This is indicated by the fact that the mean percentage of properties occupied by women exceeds the median.
45 Under the rules of coverture, the property of a married woman would be managed by her husband and listed with the rest of his property (see note Footnote3).
46 Variation is measured by the inter-quartile range (, final column) and the higher proportion of women owning property in the larger settlements (defined by the number of properties) is indicated by the fact that mean proportion in most counties exceeds the median.
47 This occurred in the Cambridgeshire parish of Over in the nineteenth century where the number of properties assessed increased more than fourfold (from 122 in 1803 to 536 in 1851) while the number of households did not even double (increasing only from 144 to 244).