Rural poor relief and agrarian capitalism

The assertion of a causal connection between the expansion of poor relief and the development of capitalist relations of production has a long pedigree in historical literature. Marxist perspectives in the 1970s, such as developed most elaborately by Catharina Lis and Hugo Soly, argued how the expansion of social policy and poor relief provisions in preindustrial Europe were to be understood as a corollary to the spread of wage labour. As the rise of wage dependence was an unsteady societal transition characterized by strong discrepancies in the demand and supply of labour, the growth of relief provisions was a necessary means of social stabilisation in a context where a growing number of people had no alternative means of support when unable to find or perform waged work. Conversely, the spread of wage labour also heightened employers’ interests in regulating an adequate supply of labour. Perceived ability and willingness to work became major determinants of relief eligibility, in order not to stimulate wilful ‘idleness’. While the ‘undeserving’ poor – those considered unwilling to work – were barred from relief and subject to repressive policies, relief provisions expanded for the ‘deserving’ poor – those considered unable and/or demonstrably willing to work. In a context of proletarianization, therefore, poor relief increasingly functioned as an instrument of elite interests regarding social order and labour regulation, seeking to check its societally most destabilising effects while stimulating the supply of labour – mediated by religious and cultural repertoires of legitimization.Footnote ²

A more recent reformulation of this argument, with an exclusively rural perspective this time, was made by Larry Patriquin, who argued that the early expansion of agrarian capitalism and the associated need for alternative support as people lost access to land, was the driving force behind England’s precocious Poor Law system in the early modern period.Footnote ³ Focused less on causality, but highlighting employer interests, was Boyer’s influential analysis how relief hand-outs at the end of the Old Poor Law functioned as de facto wage subsidies to capitalist farmers in England’s southeast.Footnote ⁴ Other researchers inversed the causality in the underlying argument, by arguing that the early development of a comprehensive poor relief system in preindustrial England was what stimulated the country’s precocious transition to wage labour, rather than the other way round: because poor relief provided some kind of insurance against the vulnerabilities of wage dependence, its existence facilitated individual decisions in favour of waged work.Footnote ⁵

In either causal perspective, the focus lay on England as a forerunner in public relief provisions. The image that the Elizabethan Poor Law of 1601 installed a precocious and relatively uniform, comprehensive, and generous system of public poor relief, which contrasted with more paltry, haphazard, and primarily urban practices of continental poor relief, long dominated historiography.Footnote ⁶ This purported contrast between England and the Continent has been subject to growing criticism in recent studies. On the one hand, English historians have demonstrated the existence of many regional, local, and temporal variations in the supposedly uniform system of poor relief.Footnote ⁷ On the other hand, continental historians have demonstrated the existence of relief practices that displayed many similarities to the English system, even in the countryside.Footnote ⁸ To the extent that the most ‘advanced’ relief practices have been observed primarily in regions characterized by agrarian capitalism, both in England and on the Continent, these observations strengthen the case for the underlying causality to be reversed again, and to privilege regional social and economic characteristics over national normative frameworks to explain variations in preindustrial poor relief.

This study revisits the notion of agrarian capitalism as a causal factor in the development of preindustrial rural poor relief by means of a bottom-up approach that starts from local data to investigate broader regional patterns from a comparative perspective. Its focus is on one rural district in early eighteenth-century Flanders characterized by the same institutional and legislative framework, but by important differentiation in terms of agricultural relations of production: one set of villages oriented towards agrarian capitalism, and another characterized by a more diversified social structure and less dependence on wage labour. The extent to which they displayed distinct patterns in poor relief, provides a way to evaluate the influence of agrarian modes of production on welfare practices. Exceptional is that we have recourse to systematic cross-sectional local data for a relatively early date, which we can compare with extant analyses for later periods and for other areas to widen the relevance of our findings.

Sources and context

The 1700 survey on poor relief was undertaken in the district of Furnes (kasselrij Veurne), situated in the north-west corner of present-day Belgium and bordering northern France and the North Sea. The district was composed of some 40 villages and governed by the same college of aldermen but was far from a homogenous social and economic space. In the northern villages, soil consisted predominantly of clay. Like in other parts of coastal Flanders, villages in the polder area had undergone a process of farm engrossment since the late Middle Ages. From the fourteenth century onwards, land was bought by urban investors and small farmsteads gradually gave way to large leasehold farms. By the late fifteenth century, leaseholders had overtaken owner-occupiers as the dominant farming group in these villages.Footnote ⁹ As a result, farms in these polder villages were notably larger than in the more inland villages located in the sandy loam belt, and social inequality was greater.Footnote ¹⁰ This differentiation in landholding structures was already apparent from the 1569 taxation registers (penningkohieren) from which we can derive the distribution of farm size. These show that farms larger than 20 ha occupied an average of 61 per cent of cultivated soil in the polder villages, but only 20 per cent in the more inland sandy loam villages, where middle-sized farms between 5 and 20 ha were dominant, occupying on average 70 per cent of cultivated soil (Graph 1). As farms larger than 20 ha were too large to exploit with family labour, their importance is an indirect indication of the prevalence of waged work by servants and labourers in the polder areas.Footnote ¹¹

Graph 1. Share of cultivated soil by farm size category in polder and sandy loam villages in the Furnes region, 1569.

Source: P. Vandewalle, De geschiedenis van de landbouw in de kasselrij Veurne, 1550–1645 (Unpublished PhD Ghent University, 1979) vol. 2., appendix 19 (N = 28)

As this process of farm engrossment continued in the later sixteenth and seventeenth centuries, early eighteenth-century polder villages were characterized by a highly polarized social and economic structure. Middling farms, relying exclusively on family labour, were nearly absent. Most of the soil was occupied by large tenant holdings, on which labour was supplied by living-in servants and local labourers, in addition to substantial numbers of seasonal migrant workers during harvest. Labour relations became increasingly oriented towards adult male day labourers and resembled those in England’s quintessential area of capitalist farming in the southeast.Footnote ¹² In the sandy loam villages, farms were smaller and relied less on waged labour. Here we find a more diversified social structure, with middling family farms occupying a central role in the economic landscape. Maintaining small plots of land and having access to more diverse income-pooling opportunities, households at the lower end of the social ladder in these inland villages combined subsistence agriculture with various by-employments to survive, among which seasonal harvest labour on large leasehold farms in the polder villages.Footnote ¹³ Within the same political and administrative district of Furnes, then, two distinct economic regions emerged from the late middle ages onwards, characterized by distinct gradations of farm engrossment and wage dependency. In this article, we aim to assess to what extent these differences in economic and social structures correlated with local variations in the organization of poor relief. To gain insight into this question, we will examine the level and composition of income and expenditure, coupled with an analysis of the number and profiles of relief recipients.

Like in other areas in the County of Flanders, public welfare in the district of Furnes was organized around parochial ‘poor tables’ (armendissen), which had been established in the late Middle Ages. The ‘poor tables’ were local institutions that controlled and managed all charitable bequests made to the local poor. For their income, they relied on donations by village members, which could consist of land, rights to land, annuities, cash, or goods in kind, of which the proceeds were used to relieve the poor.Footnote ¹⁴ In contrast to England and at least until the middle of the eighteenth century (when their income became supplemented with poor taxes), the lion’s share of poor tables’ expenses in Flanders was generated through the returns on their capital basis, in turn composed of cumulative charitable bequests from the past.Footnote ¹⁵ Earlier research for the late eighteenth century established that poor relief in coastal villages was better developed and endowed than in more inland areas, and in practice bore many similarities to the operation of the English Poor Law – although overall relief expenses were still lower than across the Channel.Footnote ¹⁶ The data used in this article allow us to venture further back in time and explore the situation at the close of the seventeenth century, when capitalistic relations of production were already in full swing in the polder area.

Our analysis is based primarily on the returns of 33 villages in the district of Furnes to a questionnaire circulated by the intendant of northern France in September 1700.Footnote ¹⁷ The questionnaire featured in plans by the French rulers to restructure poor relief provisioning in these territories recently seized by Louis XIV from the Spanish Habsburgs in the Devolution Wars (1667–1668) – an annexation consolidated (temporarily) with the Treaties of Aachen (1668) and Ryswick (1697). The questionnaire requested information about the main sources of income and expenditure of poor tables and the number and type of poor receiving parish assistance. The prime goal of the survey was to establish an overview of existing rural poor relief resources with an eye to a potential centralization of their assets for the erection of a French-style general hospital (hôpital général) in the city of Furnes.Footnote ¹⁸ From the replies to the questionnaire we can infer that local authorities were aware of the goal of this survey, which eventually was never realized. In general, they disapproved of the idea of a transfer of these resources to a central hospital and argued against the plans to deprive them of their local assets. However, as relief expenses were monitored and controlled on a yearly basis, it is unlikely that they could have supplied incorrect data.Footnote ¹⁹

To contextualise the data derived from the questionnaire, we complemented them with data from other sources relating to the 33 villages covered by the poor relief survey. These include the surface of each village, value of the land tax, number of abandoned farms, and total population for 1688 and 1697.Footnote ²⁰ From the population census of 1697, we additionally inferred information on the number of adults and children.Footnote ²¹ Lastly, we also identified the dominant soil type of each village, resulting in a subgroup of 16 polder villages in the northern part and 17 sandy loam villages in the southern part of the district (see Map 1).Footnote ²²

Map 1. Situation of the polder (16) and sandy loam villages (17) in the rural district of Furnes with preserved replies to the 1700 poor relief survey.

The survey on poor relief was organized in a short period when the region was not suffering from warfare. In 1697 hostilities resulting from the Nine Years War (1688–1697) had ceased. The region had some time to recover until it would again be the scene of military upheaval during the Spanish War of Succession (1701–1714). The survey therefore does not present a picture of poor relief during war, but during a time of peace after a long period of warfare. In 1700, this region was still recovering from the Nine Years War. Its effects on the population can be reconstructed in some detail via the population censuses that were carried out at the onset and close of this conflict.Footnote ²³ With all 33 villages studied in this article suffering population loss, the total population declined by some 32 per cent between the censuses of 1688 and 1697 (Table 1). In the polder villages total population dropped by 26 per cent, in the sandy loam region by 34 per cent. The decline in population was the result of both mortality and outmigration, as the region suffered from military disruption, multiple harvest failures, flooding, and food shortages during the conflict.Footnote ²⁴

Table 1. Population and abandoned farmsteads in polder and sandy loam villages, 1688–1697

Source: Dalle, De Bevolking (population) and City Archives Furnes, Oud Archief, nr. 320 (abandoned farms).

In addition to the destructive effects on population, warfare during this period also had an impact on agrarian relations of production. In the short term, tenants abandoned their holdings as they were no longer able to pay rent. A survey undertaken in the 1750s shows that in the course of the Nine Years War and Spanish War of Succession, a total of 167 farmsteads in this district had been abandoned, either completely or partially.Footnote ²⁵ The abandonment of farms was a feature of the polder villages in particular: two-thirds of all abandoned farmsteads were situated in this region. In the polder region, 2.3 farmsteads were abandoned per 100 inhabitants, and 0.79 per 100 ha. The abandonment of farms in the polder region illustrates the impact of warfare on economic infrastructure. Only the houses and farm buildings were vacated in the case of abandonment: the land of these holdings was typically incorporated into larger, existing holdings. The number of abandoned farmsteads is therefore indicative of the extent to which warfare contributed to the engrossment of large farms, typically at the detriment of medium-sized farms, further accelerating processes of social and economic polarization in the polder villages.

A last indicator of the situation circa 1700 is the evolution of land rent. Graph 2 shows the rental income of landed property belonging to village institutions (churches and poor tables) in the coastal villages of Slijpe and Gijverinkhove between 1650 and 1770.Footnote ²⁶ Rents started out high in the mid-century and then steadily declined to reach a low point at the close of the seventeenth century. Recovery was slow and gradual. The nominal rental value would not reach the levels of the mid-seventeenth century until c. 1765–1770. The evolution of rental income illustrates the deep economic malaise that gripped the war-affected region during the last decades of the seventeenth century.Footnote ²⁷ The low rental values of landed property not only illustrate the general economic malaise but also show that individuals and institutions who depended on land rent for their income experienced a serious contraction of their financial base.

Graph 2. Rental value of agricultural land in Slijpe and Gijverinkhove (£ per ha), 1650–1770.

Source: Calculated from G. Dalle, ‘Pachtprijzen in Veurne-Ambacht en in het Brugse Vrije’, in Dokumenten voor de Geschiedenis van Prijzen en Lonen In Vlaanderen en Brabant (Brugge: de Tempel, 1959) vol. 1: 205–238.

Relief income

As income for poor tables derived from immovable properties probably stood at a historical low in this period, the survey of 1700 informs us on the operation of a poor relief infrastructure heavily scarred by the effects of warfare and trying to recover from the heavy demographic and economic losses sustained during the period 1688–1697.Footnote ²⁸ Some villages stated in the survey that some of their lands yielded no income because they were inundated. Others complained that some of their assets yielded low rental income compared to earlier years.Footnote ²⁹ The specific goal of the survey, i.e. an overview of assets with an eye to possible centralization, also implies that the information on income and assets pertains to transferable assets only: villages were asked to list the income they derived from real estate (houses, farmsteads, and land) and (redeemable and unredeemable) annuities. In other words, only assets that were backed by a secure legal title were included in the survey. Income from poor boxes in churches and inns, church collections, small donations in money or in kind, and other ‘irregular’ sources of revenue that could not be legally enforced, were not listed – yet all indications are that such ‘irregular’ income was very modest compared to ‘regular’ sources of income.Footnote ³⁰ The survey of 1700, then, does not list all sources of income collected and distributed to the poor and most likely underestimates overall income in more steady years. However, this does not render it useless to study income patterns: because the ‘regular’ income sources reported in the survey are the cumulative result of centuries of charitable bequests, the survey of 1700 is indicative of structural local variations in the nature and level of income of these poor tables.Footnote ³¹

At first sight, sandy loam villages were on average richer in terms of welfare resources compared to the polder villages (Graph 3). Sandy loam villages disposed an average of 372 guilders (fl.) of income (median 331 fl.), while polder villages recorded an average income of 289 fl. (median 248 fl.). Yet when population numbers are taken into account, the picture changes significantly, as sandy loam villages were on average twice as large (507 inhabitants) than the sparsely populated polder villages (222 inhabitants). In the polder villages, poor tables on average disposed of 1.90 fl. per inhabitant (median 1.60 fl.). In the sandy loam villages this was markedly lower at 0.83 fl. per inhabitant (median 0.68 fl.). Relative to population, polder villages on average thus had more than twice as much poor relief income. Even when considering that wages were 10–20% higher in the polder villages,Footnote ³² the differences were marked.

Graph 3. Poor relief income in polder and sandy loam villages, 1700 (fl.).

Source: City Archives Furnes, Oud Archief, nr. 1120: Poor Relief Survey 1700 (N = 32).

In terms of wage equivalents, the average annual relief income per inhabitant in the Furnes area was relatively high: at least more than two days’ wages for a male agricultural labourer for the whole sample and at least three day’s wages in the polder villages.Footnote ³³ This was possibly slightly less than a century later, when the equivalent in annual relief income per inhabitant for the whole Furnes area in 1807 was roughly three days’ wages.Footnote ³⁴ At the same time, the 1700 figures approximate existing estimates of rate-based relief expenditure per capita in contemporary England, which was the equivalent of c. 1.7 day’s winter wages of a male agricultural labourer in 1696.Footnote ³⁵ While the per capita figures may have been boosted by the recent decline in population and there are many caveats to the comparison,Footnote ³⁶ this provides a first indication that poor relief provisions in the Furnes area were not wide off the English mark around 1700, something we will elaborate below.

Interestingly, there was a marked correlation between relief income and fiscal capacity in our sample. Land taxes in Flanders were allocated to villages following the so-called ‘Transport of Flanders’. This was a repartition scale based on the fiscal capacities of each village. To assess this fiscal capacity, assessors looked at population, land rent, and the number of farms to assess a village’s overall wealth. The relative share of each parish in this repartition table thus represents the fiscal capacity of each village.Footnote ³⁷ When we compare data on fiscal capacity on the one hand and income of poor tables relative to population on the other hand, we find a marked, positive correlation (Graph 4) ( $\hat \beta $ =11.7, p<0.001). Clearly, in relation to their population, most polder villages were characterized by both relatively high relief income and relatively high shares in the land tax – an indication of relative wealth. While retaining only around 29 per cent of the district population, polder villages held 48 per cent of all poor relief resources and contributed 50 per cent to the land tax.

Graph 4. Relief income per capita in relation to land tax per capita, 1700 (in fl.)

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700) and nr. 342 (land tax 1631).

The income structure of poor tables shows only minor differences according to soil type in our sample (Table 2). In both types of villages, the majority of income was generated from rental income of land (69% and 62% in polder and sandy loam villages respectively), while rental income from houses and small cottages accounted for on average 7 per cent and 6 per cent respectively,Footnote ³⁸ and income from annuities – which could have been donated directly or purchased by the poor table as part of a capital investment strategy – for 26 per cent and 33 per cent respectively.Footnote ³⁹ The income of poor tables was therefore determined by its capital base, in the form of (rights to) land and (rent from) monetary investments, which in turn accumulated from past donations. Although this implies an element of inelasticity and contingency, the marked correlation with the land tax suggests that the level of these donations was also governed by local wealth.Footnote ⁴⁰

Table 2. Main components of relief income, 1700

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700). Unweighted averages of respective shares per village.

Expenditure

Based on the information listed in the surveys, expenditure can be broken down into four main categories (Table 3).Footnote ⁴¹ Given the timing of our source, expenditure for maintenance and repairs of immovable properties (11%) was probably exceptionally high. Wartime destruction and abandonment of farmsteads and land probably resulted in higher levels of investment to reintroduce properties to the lease market. On average maintenance and repair costs represented the equivalent of 14 per cent of rental income from land and houses. In years and periods without war, maintenance and repairs probably accounted for a smaller share of expenditure.Footnote ⁴² In both areas, annuities (‘cijnzen’ and ‘renten’) – typically the payment of feudal or other rights attached to land that had been donated to the poor table – occupied a minor role (2%) in expenditure.

Table 3. Main categories of relief expenditure, 1700 (in fl.)

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700). Unweighted averages of respective shares per village.

Note: ‘Poor relief adjusted’ assumes that 25% of expenses towards foundations ended up as some kind of relief to the poor. See discussion in text.

A substantial share of expenses (36%) was allocated towards foundations. These arose when donations to the poor table, especially testamentary bequests, were burdened with specific, mainly religious, demands that implied recurrent expenses. For example, an individual could donate a parcel of land to the poor table and request that part of the income of this land be used to celebrate a regular remembrance mass (jaargetijde). This meant that part of the proceeds of the donation had to be reserved for the priest and other church officials to organize these masses. Whatever remained, was then used to distribute to the poor, sometimes also according to specific instructions of the donor (e.g. on specific days, to specific categories, etc.). In contrast to other types of income, income derived from so-called foundations could therefore not be freely spent according to the will of the poor relief administration but was earmarked for specific – largely religious – expenses attached to the original gift. From a detailed survey of the village of Loker we know that most of the income from such foundations was used to pay local clergy and church officials for their services: in 1700, 74 per cent of income from charitable foundations was allocated to the priest and various church officials to organize masses of remembrance. Only 24 per cent of the income was used to support the Loker poor by means of bread distributions during or after these masses.Footnote ⁴³

The importance of foundations is a reminder of the religious and symbolic meanings that imbued donations to the poor, and of their intertwinement with cultural conceptions on the relations between rich and poor. The very different degrees of involvement with foundations in the two regions therefore hint at substantial differences in the nature and functions of charitable donations to the poor tables. Polder villages on average spent almost half (44%) of their resources on foundations, to the equivalent of 0.77 fl. per inhabitant. In sandy loam villages expenditure related to foundations was only 26%, or 0.20 fl. per inhabitant. Seven out of 16 polder villages (44%) even recorded higher expenses towards foundations than for poor relief proper, while this was the case for only 3 out of 14 sandy loam villages (21%). In polder villages, against each 10 fl. of income generated from immovable property stood an average of 6.0 fl. of expenses for foundations. In the sandy loam villages, this relative ‘burden’ of foundations was only 4.5 fl. per 10 fl. of income from land and houses.

Expenditure to commemorate benefactors had an important impact on poor relief expenditure in two ways. First, charitable trusts stipulated in detail how income was to be allocated, also with respect to the material assistance provided for the poor. In most cases, remembrance masses were followed by distributions of specified goods (in most cases bread) to the poor at fixed dates. Therefore, both with respect to the timing and nature of distribution, foundations imposed restrictions on local overseers of the poor as to when and how they could spend their income. Second, higher expenditure on foundations went hand in hand with lower shares of poor relief proper in overall expenses, i.e. the combined sum of material support (food, cash, rental subsidies, boarding fees, medical assistance, burial costs) directly distributed to the poor by the table administrators: the sandy loam villages on average spent 57 per cent of their expenses on actual relief, while this was only 46 per cent in the polder villages.Footnote ⁴⁴

The lower share of income reserved for poor relief proper, however, did not imply lower actual expenses for poor relief. Even though polder villages allocated 44 per cent of their expenses to foundations, this still left them with 0.95 fl. per inhabitant to spend on poor relief proper, as against 0.46 fl. in the sandy loam villages – i.e. twice as much. If we assume that c. 25 per cent of expenses towards foundations eventually ended up as some kind of relief to the poor (mainly in bread distributions), the gap widens even further: total expenditure towards poor relief (‘poor relief adjusted’ in Table 4) was then around 1.14 fl. per inhabitant in the polder villages, but only 0.51 fl. in the sandy loam villages. Although a lower share of overall income was used for poor relief in the polder villages, then, its absolute value per inhabitant was still substantially higher than in the sandy loam villages – thanks to the overall higher income levels of relief institutions in the polder villages.

Table 4. Demographic categories of poor relief recipients, 1700 (%)

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700) and Dalle, ‘De Volkstelling’ (population census 1697).

In terms of wage equivalents, average annual expenses on poor relief per inhabitant (including 25% of expenses towards foundations) stood at 1.9 days’ wages for an agricultural labourer in the polder villages and 0.85 day’s wages in the sandy loam villages. This brings the levels of poor relief expenses in the polder villages to the average of rate-based expenditure observed for England in 1696.Footnote ⁴⁵ The Furnes per capita average may have been inflated by recent population decline, yet it was also negatively affected by a war-induced drop in revenue. Although this comparison is unavoidably heavily influenced by the wage estimates and population figures used, and complicated by different measures of endowed charity, it does suggest that the poor tables in the Furnes polder villages – even following a period of intensive warfare destruction – expended relief in the same order of magnitude as the rate-based system across the Channel at the close of the seventeenth century.Footnote ⁴⁶ This is confirmed by calculations in wheat equivalents, which yield a strikingly similar average expense of c. 12 litres of wheat per head of population.Footnote ⁴⁷

Relief recipients

Finally, we turn our attention to the poor who received parish support. The survey enquired both after the number of individuals relieved, and their subdivision according to sex between children and adults, the latter further differentiated between ‘valid’ and ‘invalid’ poor. Although none of the categories is defined or explained in the sources, we assume that the difference between both categories is whether people were deemed physically or mentally able to work, with the group of ‘invalides’ likely made up predominantly of elderly men and women. A major drawback of the data is that relief recipients were counted individually and not on a household basis, and that they do not distinguish types or levels of relief per category – yet they nevertheless allow for some interesting observations (Table 4).

In the polder villages, an average of 9% of the population received relief from the poor table. In the more densely populated sandy loam villages, this was 12%. This is more or less comparable to the situation a century later when an average of 8% in these villages received relief in 1807 (7% and 9% respectively) – and when around 10% of the English population received relief in 1802/3.Footnote ⁴⁸ On average, children made out the majority of these relief recipients (52%), while able-bodied adults accounted for 24 per cent and disabled adults (‘invalides’) for another 24 per cent (Table 4). Children made up a larger share of relief recipients in the sandy loam villages than in the polder villages: on average 58 per cent versus 46 per cent, while adults – women in particular – were more prominent as relief recipients in the polder villages.Footnote ⁴⁹ Although the interpretation of these figures is hampered by a lack of insight into household contexts and types of relief, and by small numbers of recipients in some villages, they do at the very least indicate that relief to able-bodied adults was not uncommon.

We can put these figures in further perspective by relating the number of relief recipients per village to the 1697 census returns, which also allow us to differentiate between children and adults of both sexes, thus providing an indication of the relative likelihood of being relieved for children, men and women respectively (Graph 5).Footnote ⁵⁰ It confirms that children were considerably more likely (18%) than adults (7%) to figure on relief lists relative to their number – especially so in the sandy loam villages, where no less than an average 23 per cent of village children received relief, as against 13 per cent in the polder villages. Because most of these children likely lived with their families, relief to children probably also amounted to indirect relief to their parent(s). As to the likelihood of adults to be relieved in their own right, divergence between village types was smaller, and that between the sexes universal, with women (9%) on average more likely to receive relief than men (6%). The overall predominance of children and to a lesser extent women on the Furnes relief lists is no doubt indicative of their greater likelihood to be considered deserving, but possibly also a reflection of – at least in the case of women compared to men – lesser income opportunities.Footnote ⁵¹

Graph 5. Share of relief recipients per demographic subgroup, 1700 (%).

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700) and Dalle, ‘De Volkstelling’ (population census 1697). Unweighted averages of respective shares per village. Shares calculated as number of child and male/female adult relief recipients in relation to number of children and male/female adults in the 1697 census returns.

Although providing relief to able-bodied men is often identified as one of the ways in which the Old Poor Law was used by large farmers in England to subsidize wages, this probably happened only in the last years of the eighteenth century and early decades of the nineteenth century on any systematic scale and was probably not as widespread as sometimes assumed.Footnote ⁵² Recent micro-research has in any case demonstrated that myriad variations existed and that the bulk of English relief recipients even in the late eighteenth century consisted of classic ‘deserving’ poor: elderly, sick, women and children.Footnote ⁵³ Although the picture is hampered by their ‘atomised’ representation, the overall composition of poor relief recipients according to the 1700 poor relief survey for the Furnes district, then, does not strike as necessarily very different from the profiles of recipients across the Channel – insofar as any generalizations are warranted.

Important differences in distribution practices between the two areas in the Furnes district pertained to relief expenses per recipient (Graph 6): paupers received on average 14.5 fl. per recipient in the polder villages, versus only 5.9 fl. in the sandy loam villages – not taking into account relief awarded via foundations.Footnote ⁵⁴ This was the equivalent of 24 and 10 day’s wages of an agricultural male labourer respectively, and of 160 and 65 litres of wheat.Footnote ⁵⁵ Moreover, the variance was much more restricted in the sandy loam villages than in the polder villages, where 4 out of 16 villages spent more than 27 fl. per recipient, i.e. more than five times the average of the sandy loam villages. These marked discrepancies clearly implied substantial differences in the value, function, and meaning of poor relief in both village types. Unfortunately, the data do not allow us to differentiate average expenses per recipient subgroup.

Graph 6. Relief expenses per recipient (fl.), 1700.

Source: City Archives Furnes, Oud Archief, nr. 1120: (poor relief survey 1700). Excluding relief from foundations.

Although more relief was available and spent in the polder villages, then, it was distributed among a smaller share of its population than in the sandy loam villages, where poor relief was both less selective and less substantial. Whereas the sandy loam villages appear to have relieved more families in a less discriminate manner, relief in the polder villages appears to have been biased more towards female-headed households – although able-bodied males did figure on relief lists. Both in sandy loam and polder villages, then, poor relief was an important vector of local society in 1700 – by implicating a large share of the population and/or providing substantial relief. Even in the sandy loam villages, average relief per recipient was the equivalent of at least ten day’s wages of a male agricultural labourer, and its wide reach is illustrated by the fact that almost one in four children was in contact with the poor relief table at any given point in time. Although fewer people were relieved in the polder villages, their share still amounted to almost one in ten, and they received the average equivalent of approximately a monthly wage of a male agricultural labourer. Considering that these data present a cross-sectional snapshot and that relief dependence varied over one’s lifetime, it is not unlikely that a large share or even majority of the population was relieved at some point in their lives.Footnote ⁵⁶ Although differences regarding the demographic background of relief recipients are obscured by lack of detail in the survey, the few indications we have, hint at different levels of dependency in the two regions. Unfortunately, the static nature of the underlying data does not allow us to retrace how dependency unfolded over the life course.

Conspicuous charity and agrarian capitalism

The above analysis of the 1700 survey on poor relief has demonstrated that polder villages in the northern clay area of the Furnes district, which were characterized by large holdings operated by wage labour, recorded significantly higher relief income relative to their population than the more inland villages in the sandy loam area, where middle-sized family farms were more important. With a smaller share of the population receiving relief, this resulted in markedly higher relief levels per recipient in the northern villages. These differences are consistent with spatial patterns observed later in the eighteenth and early nineteenth centuries, and confirm a positive correlation between spending on rural poor relief and the prevalence of agrarian capitalism.Footnote ⁵⁷ As these villages all belonged to the same administrative region with the same institutional framework for poor relief, these observations strengthen the argument that wage dependence promoted the expansion of public poor relief rather than the other way round.

To evaluate the ways in which poor relief and agrarian capitalism interacted in these villages, it is useful to consider dynamics of affordability, necessity, and legitimacy. One observation is indeed that northern villages were also richer, at least as measured by land tax revenue, which reflects the higher productivity of large capital-intensive farms, and by extension, of agrarian capitalism. Yet it does not necessarily follow that polder villages allocated more resources to poor relief simply because they could better afford it: the specificities of the processes of both receiving and giving aid in these villages indicate more complex causalities at work, interacting with necessity and legitimacy.

The greater reliance on wage labour, so we argue, indeed created specific needs in terms of poor relief in the polder villages. This was not so much because of wider poverty per se: with relatively higher wages, overall conditions for healthy adult male labourers during the busy season were no doubt relatively comfortable. As other studies have highlighted, however, the main risks of rural proletarianization lay in the absence of other, independent sources of income, especially for women and children whose income opportunities in agrarian wage labour were considerably lower, to rely on when sickness, old age, family abandonment, widowhood, (seasonal) unemployment or childbirth upset wage-dependent households’ precarious balance between income and needs.Footnote ⁵⁸ In this context, it is worth reminding that much wage labour in the polder villages was temporary, supplied by male seasonal migrants, whose household income pooling typically included the cultivation of small plots of lands at home.Footnote ⁵⁹ Local labouring families, however, had no other recourse than public poor relief when misfortune, lifecycle, or seasonal unemployment deprived them of their livelihood. In the sandy loam villages, poverty was also rife, but not as ‘deep’ as in the polder area. This was because opportunities for income pooling were more diverse in the sandy loam villages and social polarization was less pronounced: here, mixed landholding structures offered greater access to land, by-employments, and informal support networks embedded in ‘local economies of exchange’ – thus lessening overall dependence on public poor relief to cope with personal vulnerability.Footnote ⁶⁰

Unfortunately, no direct measures of poverty and need in these villages exist to support the argument, apart from those derived from poor relief accounts. The only additional, convoluted, indication is the share of inhabitants exempted for reasons of poverty from the milling tax in 1688, a modest head tax in the area, which averaged 14% of residents in polder villages and 11% in sandy loam villages.Footnote ⁶¹ Of course, the criteria of fiscal poverty are complex to interpret, and the Nine Years War separates the fiscal data from the relief survey.Footnote ⁶² Yet, the observation that the average share of fiscal poor exceeded that of relief recipients in the polder villages, while the reverse was true in the sandy loam villages, in any case, fits with the argument that the degree of necessity among relief recipients was larger in polder villages than in inland villages, as was that of selectivity in determining relief eligibility.

This distinction between ‘deep’ and ‘shallow’ dependency helps to explain why relief was both more concentrated and more substantial in polder villages, and less selective and less generous in the sandy loam villages, where many were no doubt poor, but also had more access to additional resources to make ends meet. The greater restrictiveness in the northern villages also resonates with the older argument in the literature that associates increasing selectivity in the definition of deservingness – primarily in terms of ability and willingness to work – with the expansion of wage labour, in order to ‘activate’ workers.Footnote ⁶³ Restricting relief to vulnerable and invalid poor, hindered its use as a direct wage subsidy in ways envisaged by Boyer.Footnote ⁶⁴ Yet it still amounted to an indirect subsidy of the cost of labour, and may have been more relevant when much labour was supplied by seasonal migrants anyway.Footnote ⁶⁵ The greater prevalence of women among recipients, was likely both a reflection of their vulnerability and of a greater concern not to reduce work incentives for able-bodied male workers. Conversely, as in the English southeast, local farmers may have supported the allocation of relief to cover temporary income squeezes in their workers’ households, in order to maintain a minimal labour supply in situ.

One additional argument that polder villages did not provide more assistance than inland villages simply because they were better endowed, but on the contrary were under pressure to meet local needs, is that they regularly experimented with additional ways to raise money for the poor table, and would be among the first to integrate regular poor taxes in their revenue strategies when this was facilitated in the region some fifty years later.Footnote ⁶⁶ Although northern villages had considerably more income to spend on poor relief, then, spending was not blanket nor distribution indiscriminate. These reflected, we believe, the specific vulnerabilities associated with rural wage dependence on the one hand, and cultural repertoires of deservingness that activated male wage labour and legitimized heavily polarized social relations on the other hand. This is also reflected in the specific labour legislation that characterized this region. In the northern villages labour legislation was much more coercive during the early modern period – including compulsory service – and receipt of poor relief was dependent on compliance with the stringent labour laws.Footnote ⁶⁷

That brings us to the dimension of legitimacy. In polder villages, donations were more heavily burdened with masses of remembrance than in the sandy loam villages. Benefactors in the polder region appear to have attached considerable importance to the symbolic and conspicuous dimension of charitable donations. This ties in intelligibly with research on local political culture in coastal Flanders: here, village institutions, such as the poor table, played an important role in the ambitions of village elites to establish and consolidate the prestige of family lineages and to reproduce themselves culturally. As jaargetijden served to keep the memory of charitable actions by rich families alive, they formed part and parcel of local strategies of cultural representation and legitimization by polder elites. These strategies were at least partly fostered by the high levels of inequality in these villages, which made local elites both more able and willing to conspicuously legitimize and demonstrate their privileged position by lavish foundations.Footnote ⁶⁸ The argument is not so much that donations were necessarily instrumental to elite interests in any direct or conscious way, but that they underpinned a cultural framework in which religious and charitable prestige supported the political legitimacy of local elites in a context of social polarization.

Lastly, it is useful to reflect on the chronology of the interactions involved. That income for poor relief was derived primarily from past donations, inserts an unspecified time lag in the analysis. Poor relief accounts routinely do not date the origins of their endowments and have not been conserved in a continuous way to allow for easy reconstruction. Some retrospective insight can be gained, however, from a survey of extant foundations drafted in 1770 for villages in the Liberty of Bruges, an adjacent region of coastal Flanders.Footnote ⁶⁹ The earliest foundations on this list originated from the sixteenth and sometimes even fifteenth centuries, but they were generally surpassed in number by those dating from the seventeenth and, to a lesser extent, eighteenth centuries. Although many were also undated, the survey at the very least demonstrates that foundations with poor tables were still very much en vogue at the end of the seventeenth century in coastal Flanders.Footnote ⁷⁰ That some sources of income used by the poor tables already had a venerable history, needs moreover not to undermine the posited causal chronology. Processes towards farm engrossment have been present in the polder areas since at least the late Middle Ages, and by the sixteenth century, agrarian capitalism was already well-developed here. While some foundations may have had a venerable history when reported as a source of income in 1700, they originated from a local context characterized by the gradual development of agrarian capitalism and reflected an enduring tradition that continued to be replenished by new donations and foundations.