1. Introduction
Land transmissions were a crucial factor in the socio-economic reproduction of pre-industrial societies and in the development of land markets.Footnote 1 The farm occupancy system and land transfers were closely linked to social and geographical mobility. This article addresses how long tenant farmers in the Russian Baltic Provinces of Estland and Livland managed to occupy the farms and whether they transferred them within the family in the early phase of agricultural transformation (1841–1889).
Analysis of the changes in farm occupancy in Estland and Livland contributes to the long-standing debate over the relative power of manorial lords and tenants in the (East Elbian) manorial system. This debate constitutes part of a greater discussion over the choices of the peasants and the constraints of those choices in the pre-industrial countryside.Footnote 2 Classical historiography stresses the subordination that characterised the relationships between the rulers and the ruled in the areas dominated by the East Elbian manorial system.Footnote 3 In the more current research, it is common to accentuate mutual dependence where both parties had reciprocal rights and commitments.Footnote 4 The peasants’ lifelong or hereditary tenure of their farms in Western Europe has traditionally been contrasted with the more insecure situation of eastern European peasants.Footnote 5 Recent research, by contrast, has questioned the absolute power of landlords in respect to land transmissions and has revealed a gap between ‘the theoretical powers of the landlords and the variable practice of lordship’ in eastern and eastern-central Europe.Footnote 6 For example, Alice Velková and Hermann Zeitlhofer maintain that in Bohemia manorial interference in land transfers was limited and landlords mainly intervened in land transmissions only when farms were seriously mismanaged.Footnote 7
In the Baltic provinces, we have, on the one hand, either labour obligations, money rent or purchase payments determined by an estate owner and, on the other hand, the reactions of tenant peasants to lords’ incentives. On noble manors, peasants were constrained by insecure tenancies: leases were short term, the estate owner could easily terminate the contract if the tenant did not comply with its terms, and tenants had no prerogative to renew the contract if they fulfilled duties and obligations. Tenants could bequeath the farm, but there is no evidence that they bought or sold any pieces of land. Looking at the reactions of farming peasants, the question arises whether the Baltic peasant families normally tried to retain the farm and relinquish it to their offspring. Andrejs Plakans and his colleagues have described how farming families in the pre-emancipation period as well as immediately after emancipation in the early nineteenth century tried to prevent themselves from being dispossessed.Footnote 8
The traditional understanding of the continuity of the farm within the family in the pre-industrial countryside has been challenged by several recent microstudies of different parts of Europe, which suggest that peasants made individual and profit-maximising choices about using land, transferring its ownership, and participating in market transactions.Footnote 9 Practices of land transmissions in the stem-family systems varied from country to country and context to context and were not time-invariant, as in some regions family farms became increasingly permanent over the course of the nineteenth century.Footnote 10 In particular, a study of Scania has mentioned the role of differences in property rights (freehold vs. tenancy) in this regard: freeholders seem to have passed holdings through the family, while tenant farmers most often relinquished their holdings to unrelated individuals.Footnote 11 Swedish crown tenants had secure and hereditary tenancies and the land could only be inherited within the family.
In the following discussion, continued occupancy of a family farm is seen as both an indicator of the continued viability of a tenant's household despite changing conditions, as well as a reflection of the manor owner's economic interests and incentives. The study aims to show whether the tenants in a zone subject to strong manorialism could keep the farm in the family and meet the demands of estate owners (noble lords as well as the crown). The article addresses the question of whether the family (not an individual) retained the farm. There was a major difference between state estates and noble estates in respect to transitions in tenantship: on state estates, the incoming tenant was expected to be a son, son-in-law or brother of the outgoing tenant; on noble manors, the landlord could rent the farms to whomever he chose. I do not merely seek to demonstrate how landowner economics determined peasant economy in Estland and Livland, but aim also to analyse the determinants of regional and local variations in the permanence of family farms. Farm occupancy systems, including evictions, are linked with localised ecological characteristics and legal, economic and social changes at the macro level. I make these connections in order to elucidate how the tenants managed to cope with three major challenges facing them in the post-emancipation period: a subsistence crisis in the mid-1840s, transition to money rent in the late 1840s–1860s and transition of farmland into freeholds in the 1850s–1880s. The topic will be discussed from the perspective of socio-economic history and the peasants’ point of view.
In the Baltic case, several elements were combined, making it difficult to determine the relative importance of legal, economic and ecological factors: tenure and inheritance rights were weak on noble land; labour dues were unlimited on noble land and fixed on state land; the shift from labour to money rents and from tenancy to freehold lasted over decades and on noble manors were dependent on the will of a landlord; and migration was limited, and household composition and size varied.Footnote 12
Comparative material has been drawn from several countries with manorial systems in which manorial land ownership was the dominant form of land ownership and the peasants were rendered dependent on the manorial lord. In many settings, the dependency took the form of serfdom, but the manorial system could persist also after the abolition of serfdom as was the case in the Baltics or without serfdom as it was in Scania. Differences in either methodologies or land use systems render comparisons mostly unfeasible. For example, in Russia proper during the time of manorialism, there were no family farms due to the communal land allocation system: households were entitled to a share of the communal land which was periodically (partially) redistributed. The rent-extraction powers of Russian landlords were not constrained by custom or law, but the question of whether Russian landlords exploited the peasants to the extent that they lost their plot of land and fell into the category of landless ‘poor serfs’ is yet to be studied.Footnote 13 The Russian partible inheritance system further complicates comparisons with Estonian practices.Footnote 14 Whereas studies on manorialism in eastern and central Europe tend to focus on the early modern era, Swedish research has also examined the period when market expansion and commercialisation promoted more active landlord strategies. As such, Swedish research offers useful comparative insights.
In line with recent studies exploring practices in manorial societies, which varied across space and over time, and from landlord to landlord, this article addresses a specific situation on a micro-level.Footnote 15 More empirical work on various measures of economic performance and well-being is necessary.Footnote 16 The permanence of farms could serve as an indicator for comparing manorial societies with each other, and with societies in which farmers were not constrained by estate owners.
After a short historiographical overview, sources, methods and institutional setting will be described. The two main sections discuss the changes of farm occupancy in the time of a subsistence crisis and in the time of transition to money rent and outright purchase of farmland.
2. Historiography
Most historians of the subject agree that in the decades immediately following serf emancipation in 1816 in Estland and in 1819 in Livland, farms often changed hands as a result of increased labour services and taxes, insecure tenancies, implementation of short-term leases and demesne expansions by landlords. They also point to substantial local differences in the permanence of farms across the Baltics.Footnote 17 Whereas transition to money rents on noble lands has received very little interest in this respect, the capacity of tenants to buy their farms outright – to transform leaseholds into freeholds – has gained much attention. On the basis of supplications and court records, Artur Vassar claimed that on noble lands, many farms changed hands at the time of the purchase of the freehold.Footnote 18 Another prominent Estonian agrarian historian, Juhan Kahk, used similar sources but came to the contrasting conclusion that ‘the vast majority’ of tenant farmers kept their holdings in those years.Footnote 19
Plakans and his colleagues have studied changes of headship in tenant households from the perspective of peasant family strategies in Piņki (Pinkenhof).Footnote 20 Plakans et al. claim that ‘under serfdom, peasants strove successfully to pass on to kin, and especially sons, the single most important position in their society. After the abolition of serfdom, they still sought to bequeath headships to kin and sons, but were not nearly as successful’.Footnote 21 In pre-emancipation Piņki, 58 per cent of land transfers went to relatives, but this share decreased to 40 per cent between 1833 and 1850. In fact, the cross-sectional data they used do not distinguish between changes of headship and changes of occupation. Their findings actually indicate that changes of tenancies became more frequent between 1833 and 1850 and this is expected given the insecure tenancies. Although an estate owned by the city of Riga (Piņki) can hardly be considered representative of other estates in Livland, the Piņki case nevertheless describes the general trend in the post-emancipation period fairly well. By examining farm and family histories in three villages on the Pakri peninsula in Estland, Tiina Peil and Madeleine Bonow have recently questioned the alleged ‘immobility’ of nineteenth-century farmers, arguing that social mobility was considerable and combined with physical movement.Footnote 22
3. Data
This study examines changes of farm occupancy and the origins of the incoming tenants in five parishes in the Estonian-language area of Livland and Estland in the period 1841–1889 (Figure 1). Although the study considers the territory of today's Estonia, I will use the historical terms Estland (northern Estonia) and northern Livland (southern Estonia) since each Baltic province had its own agrarian laws and regulations and I refer to province-level statistical data. In Ösel (the island of Saaremaa), its own noble corporation exercised power but administratively it was a county of Livland. As there is little reason to doubt the claim that farms often changed hands in the early post-emancipation period, the analysis starts with the crisis of the 1840s. It ends with the 1880s, when the process of turning farmland into freehold land was largely over in the northern Livland mainland. The system of registering land properties changed in 1889, and thereafter it becomes far more complicated to systematically follow sales of farmland by landlords.
Figure 1. Baltic Sea region, Estland and Livland in the nineteenth century. 1 – Saarde, 2 – Helme, 3 – Viljandi, 4 – Jämaja and 5 – Martna. Drawn by Anto Aasa and Kersti Siitan.
Farm histories have been composed using rent and purchase contracts,Footnote 23 census data,Footnote 24 parish member lists (Personalbücher),Footnote 25Wackenbücher (inventories of peasant obligations) and granary booksFootnote 26 that are linked to parish registers and yearly listings of in-migrants. In the Baltics, there are no consistent and comprehensive sources on land transfers – as, for example, ‘Grundbücher’ in Bohemia or manorial rentals – which makes it difficult to trace all changes of tenancy with yearly precision. Whereas written rent contracts became obligatory in Livland in 1853 and in Estland six years later, information on earlier changes in tenancies derives mostly from church records. The varying quality of data in censuses and church records and in some cases their fragmentary survival complicate the study. The practice of dropping the old farm names, as well as the dissolution and formation of farms during the land settlement campaigns and enclosures in the parish of Martna and on the estates of Olbrüki and Viljandi, pose another great obstacle to tracing farm histories. For such reasons, a few estates are excluded from the analysis, which decreases the number of farms under observation but does not change the overall proportions of stayers and leavers. Better-documented parishes include Helme and Saarde, and the least documented area is Martna.
In order to explain the micro-data on changes of farm occupancy and show the relative power or powerlessness of the two parties involved, community and parish court protocols and supplications are included in the analysis. The very scant number of preserved court records, as well as their fragmentary nature, however, allows only qualitative use. The 1850s and 1860s are considerably better covered than the 1840s. The complaints of those peasants who appealed to the district court or the Baltic governor-general are much better preserved than parish court files. Peasants could make only verbal complaints and petitions and had gone to court in person. Moreover, they had to watch what they said, as ‘baseless’ complaints against the lords that were considered to be motivated by ‘malice’ were punishable by arrest or corporal punishment.Footnote 27
The study compares the three parishes of Helme, Viljandi and Saarde in economically more advanced and prosperous northern Livland with Martna and Jämaja, both of which were environmentally as well as economically dissimilar to each other. The parishes studied, of course, cannot be considered a random sample of all 106 rural parishes in the territory of today's Estonia. However, they were typical of three different socio-economic areas – northern Livland, Estland and Ösel. In all studied parishes, agriculture played a major role in the sustenance of the population. Until the introduction of money rents and outright purchase of farms, the parishes did not differ much from each other in terms of what was grown and only a very small portion of farm products were marketed. Northern Livland (especially its southern parts) was mostly characterised by large farms and households and fertile soils. The agricultural transformation and commercialisation of production happened more quickly there than in Estland. Agricultural development in Ösel lagged behind the mainland and the island retained traditional farming practices (open-field system and common pastures, a three-field system, and subsistence farming) well into the twentieth century. Ösel farms were small and medium-sized, and supplementary income from wages of migrant labourers as well as from fishing played an important role in the sustenance of rural families. In Jämaja, low soil fertility and thin topsoil contributed to the persistence of older methods of cultivation and a subsistence mode of farming.
The population, number of manors and number of farms in each parish are presented in Table 1. Because the population figures include those not living on farms (demesne land, cottagers, etc.), we can only get a rough idea of the size of the average farm household, not included in Table 1. Viljandi town, located in the western part of Viljandi parish, had a population of 2,406 in 1862. The other parishes were totally rural. Overall, in 1863, the urban population formed 8.7 per cent of the population in Estonia.
Table 1. The number of manors, people and farmsteads in 1850a
a Data are unavailable on Ehmja and Rannamõisa (Martna). Taagepera (Helme) and the state estates of Väike-Kõpu and Välgita (Viljandi) are not included in the study. Town population is excluded.
b Includes also people living on demesne land, cottagers, etc.
4. Methods
Earlier studies about changes of farm occupancy in the Russian Baltic provinces either confine themselves to a few microstudies or refer to narrative sources. This article instead uses individual-level data and follows the change of occupancy on more than 1,000 farmsteads across several parishes from different areas of Estonia in order to avoid generalising on the basis of too small a number of cases and presenting the exceptional as typical.
Instead of tracking particular farming families, the article takes the farm as the base unit and traces the comings and goings of farmers.Footnote 28 In the rather rigid manorial system, both the interests and arbitrariness of the manorial lords and the strategies of tenant families should be considered. One might think that the interests of manorial lords and tenants often coincided, since the manor was interested in maintaining a stable labour force and secure income from well-run farms, and this would have favoured prolonging contracts. In reality, it holds true only partially, and court records, as well as supplications, reveal frequent clashes of interest. Manors produced for the market and had to be profit-making enterprises.Footnote 29 The manorial lord had much incentive to expand the estate's arable land (land farmed directly for the lord's profit with corvée or hired labour) at the expense of peasant holdings and to extract as much labour time and money rent from the tenant farmers as could be obtained without significantly damaging the ability of farm households to continue to generate profits for the manorial lord. Baltic peasant households and communities, and not manorial lords, were responsible for maintaining elderly tenants.
For most manors, the archival sources do not enable us to present accurate percentages showing how many changes in occupancy were due to evictions and financial failures, to the death of the tenant or lack of an heir, or because the occupants could no longer afford the farm. In the preserved court records and supplications, we can see that several former tenants vehemently protested their expulsions not only at the courts that were controlled by the noblemen but also sought to defend their interests against the overwhelming demands of the manorial lords by addressing the Archbishop of the Russian Orthodox Church in Riga, the head of the gendarme and the Baltic governor-general, or the Minister of Interior and the Emperor.Footnote 30 They petitioned the Emperor individually as well as collectively. Singular petitions indicate serious social conflicts when they form part of a larger body of similar petitions. Peasants’ attachment to their farms was also reflected in their rhetoric.
Not all changes of occupancy can be associated with being disturbingly in arrears to the estate or the granary, impoverishment or the termination of the contract by landlord. Several tenants sought a better or suitable farm for themselves or transacted with kin and non-kin.Footnote 31 The farming family could lose its head or a substantial number of able-bodied male labourers. Mortality also took a toll on the viability of households.
Comparisons reveal that there were systematic differences between different types of manors, different types of landlords and different parishes. Systematic differences in farm permanence between manor types (state vs. noble) and types of manorial lords (conventional vs. paternalist) point to external factors but not to internal ones (e.g. tenant's motivations). Changes of occupancy on state estates serve as a baseline against which turnover rates elsewhere can be measured. On these estates, the manor renters had few avenues to intervene in land transfers, and after the introduction of hereditary tenure, they lost even these avenues.
Former tenant families can also be tracked to see whether they settled on another farm or fell into the ranks of the landless. If the tenant family fell into the ranks of landless peasants, it may be assumed that leaving the farm was not their own choice but, most likely, a step forced upon them by their inability to meet the manorial lord's demands. Tenants’ refusal to conclude rent as well as sale contracts on the manorial lord's terms gives evidence not of their will to abandon the farms but of their unwillingness or inability to acquiesce to excessive labour dues, an increase in rent, or other unfavourable terms of the contract.Footnote 32 Differences between manors cannot be explained by demographic accidents, as these were random across the study area (in the same period of time), and the death of the family head did not necessarily lead to dispossession: on one type of manor it did, while on another, it did not. Again, dispossessions were based not solely on the occurrence of the tenant's death but on the contract terms and the manorial lord's motives.
In order to keep their tenancies and social status, tenants also resorted to other strategies in addition to protests and lawsuits. Plakans et al. have stressed the importance of the large household units, composed of kin and non-kin, as one element of tenant strategies for guaranteeing their survival and well-being when faced with the demands of landlords in the time of corvée.Footnote 33 The head of the farm kept married offspring on the farmstead so that at retirement or early death, the current head could be replaced by a married successor. Tenant farmers could follow property-related marriage strategies that brought the property into the marriage and expanded family networks which could be exploited in times of stress. An increasing and pronounced pattern of social homogamy among landed peasants in Helme suggests that strategies aimed at securing a viable landholding and social reproduction lay at the centre of the marriage decision.Footnote 34 This result is in accordance with what has been shown for various regions of Europe; children of landed peasants usually married their equals and rather strong social homogamy characterised marriages of landowning peasants.Footnote 35 This stands in sharp contrast to Peil and Bonow's findings from Harju-Madise parish in north-western Estland, where no obvious group preferences in choosing a marriage partner existed in the nineteenth century.Footnote 36 Peil and Bonow's finding, however, might hold true for Martna and Jämaja, as one might assume that marriages were socially more heterogamous if opportunities for capital accumulation were constrained due to the heavy burden of rents and dues, the small size of landholdings and environmentally unfavourable conditions for farming. Both findings can be explained by the fact that as long as the land property was insecure or commercialisation of peasant economy low, the farmer peasants did not consider the socio-economic status of the bride very important. Marriage choices became very economically rational when the outright purchase of farmland began when it became crucial to pick the right partner with a large dowry and family network.
The results of this study by no means exhaust the many possible ways in which the subject can be treated, but they are sufficient to show the relative instability of tenant holdings and lack of independence of the tenants in land transfers in the north-eastern margins of the European continent. At the family level, losing a farm can normally be interpreted as a failure. Earlier research has suggested that farmers’ incentives for long-term investments were diminished by their insecure tenancies,Footnote 37 and this general statement is supported by the evidence from Estland and Livland: the capitalist transformation of the farm economy occurred along with the transformation of tenancies into freeholds.Footnote 38
5. Institutional setting: farm occupancy system
Nineteenth-century agrarian reforms in the Russian Baltic provinces differ from similar processes in the territories of Russia proper, Poland and Prussia in one important aspect. Namely, farming peasants did not receive a title to land along with emancipation, and several decades separated the arrival of personal freedom from the purchase of land and transformation of tenancies into freeholds. Until the mid-nineteenth century, virtually all agricultural land in the Russian Baltic provinces belonged either to nobles or to the Russian crown. Cities and the Lutheran church also owned landed estates.Footnote 39 Both state estates and noble manors were large-scale agricultural enterprises. In return for a farmstead, the tenant was normally required to pay labour dues, which were based on a ‘free agreement’ between the tenant and the estate owner. On noble estates in 1816 in Estland and in 1819 in Livland, farmland was rented to peasants on an annual, three-year or six-year contract basis. The agrarian laws of 1849 (in Livland), 1856 (in Estland) and 1865 (in Ösel) extended the minimum term of the contract to six years, and in Kurland (1863) to 12 years.Footnote 40 On noble estates, corvée labour – which often had made it difficult for the tenant to carry out his own farming – was not banned until 1868. The agrarian laws also governed land sales to peasants who could buy holdings by means of long-term mortgages. Peasant land on noble manors was brought into the market through its transformation into freehold land, which started in Livland very slowly in the 1840s–1850s. Although the purchase and sale of peasant land remained within free-market conditions, tenants in Livland and Kurland were given the pre-emptive right to land purchase in 1865 and 1863, respectively, with a four-week or six-week period granted for deliberation.Footnote 41 The differential timing of reform in the three Baltic provinces also contributed to the further differentiation of the regions in terms of the pace of transition to money rent and outright purchase of farmland. Because noble landlords owned most of the land, they largely determined the duration of tenancies, rent levels and land purchase prices. Baltic tenants did not enjoy strong tenure and inheritance rights and those living on noble and church estates lacked almost any legal rights in land transfers.
The manor had great power in determining the terms of the contract, and the peasant emancipation acts rather strictly prescribed how the tenant was to cultivate the land, treat the buildings and the ‘iron inventory’ (the fixed amount of goods of each farmstead), which were the property of the estate owner.Footnote 42 The relatively small number of farmsteads per estate made keeping a careful watch over them possible.
On state estates, in contrast to noble manors, the tenants had the prerogative to renew the contract if they fulfilled duties and obligations to the estate and to the crown.Footnote 43 The maximum amount of labour dues that the leaseholders of state estates could require, as well as money rent paid to the crown, was fixed. The dismissal of tenants disturbingly in arrears or mismanaging the farm required, at least formally, consent of the community court, parish magistrate and the district inspector of state domains. The last link in this chain occasionally reversed the eviction decisions or, if a farm became vacant, ensured that it was transferred to a peasant from the same community.Footnote 44 Until money rent replaced labour rent, the leaseholder had to find a new tenant for the farm, but in practice, he did this in cooperation with the local peasant community. After the introduction of money rent paid directly to the crown, the leaseholder of the estate only asked the office in charge of state estates for its consent to the communal court's decision on the new tenant and provided the office with his opinion about the candidate. In practice, the solid and ‘reliable’ local men had the best chances to take over, but sometimes the outgoing tenant could decide his successor or his debt guarantor became the head.
Hereditary tenure of farms on crown land was established in 1869, almost a century later than in Sweden.Footnote 45 In case of eviction because of rent arrears, his heirs or relatives had the prerogative to take over the farm. From 1859, crown tenants had the right to buy their landholdings and turn them into freeholds. The 12 June 1886 law made the redemption of farms compulsory and the purchase price could be paid during the subsequent 44 years. Redemption payments on state estates were on average 2–2.5 times smaller than purchase prices on noble estates.Footnote 46 The rental levels had differed similarly between them in earlier years.
Several legal restrictions and prescriptions influenced the composition and size of the Baltic farmstead as well. Agrarian laws discouraged dividing a farm and allowed only one heir to inherit, but they did not necessarily prevent setting up cottages on common lands or on the outskirts of a farmer's allotments. The laws fixed the minimum size of peasant farm holdings on noble manors, and a new farm could be set up either on the manorial lord's initiative or with his approval. Servants, cottagers, artisans and other landless peasants, who constituted over half of the rural population, were ‘attached’ to an existing farmstead. In Estland and Ösel, however, cottagers formed separate households, and thus, the mean household size of farms was smaller (8–10 people).Footnote 47 By contrast, in Mulgimaa (e.g. Helme)Footnote 48 and in the parish of Viljandi, residential groups were numerically large and diversified with respect to kinship and generational structure and a typical farmstead in 1834–1858 contained 8 or more adults (with children) or 2–3 conjugal units.Footnote 49 Laws and decrees checking migration to towns and even to other peasant communities restricted migration. Several restrictions on freedom of movement for the peasantry were withdrawn only in the 1860s.
The principles controlling the relinquishing of property to the next generation also contributed to the household pattern. Generally, the eldest son (or a married daughter) inherited the parents’ holding, and co-successors received their portions in cash, livestock or goods.Footnote 50 If no male siblings survived, farm succession continued down the female line, but the husband was responsible for farm management.
Peasant emancipation laws and mid-century agrarian laws determined that, in the case of a tenant's death, the transaction was cancelled after the current rental year. The death of a tenant, regardless of whether he had able-bodied male descendants to take over, on some noble manors led to the family's dismissal from the farm. For example, the lord of Lõve manor sold Peebu farm to a stranger a few months after the death of the tenant who had adult sons waiting to take over the parental farm. Their lawsuit against the lord was unsuccessful.Footnote 51 However, if the rental contracts contained an inheritance clause, the farm would go to the successor. From Table 2, it is clear that a tenant's death did not have a major effect on the permanence of family farms even on noble manors, as in at least 72.5 per cent of cases the farm was passed on within the family. We might assume that in Helme parish, where the landlords’ interventions were rather common and tenancies changed hands more often than on average, a tenant's death affected the permanence of farms more than in other parishes but even in this context, the high turnover among tenants cannot be explained mostly by demographic accidents. If a household head wanted to continue managing the farm and have his descendant take over, he had to keep at least one or two adult children in the household and not let older son(s) separate from him, ensuring that upon his death the economic consequences for the family would be minimal.
Table 2. Relationship of the new farmer to the previous farmer in the parish of Helme, 1844–1867a
a Helme church estate, noble manors of Helme, Jõgeveste, Leebiku-Vanamõisa, Lõve, Patküla and Riidaja (Lutheran population).
6. The crisis of the mid-1840s
Harvest failures and grain shortages were a routine part of agriculture and socio-economic relations in the early modern era. Hunger crises, by contrast, were rather exceptional. However, in the first half of the nineteenth century, northern Livland and Estland witnessed several crises of varying severity, including one that hit the provinces in the mid-1840s. Figure 2 presents the relative differences in mortality between the study areas in the 1840s, with 1837–1839 (a non-crisis period) as the baseline. The death toll was significantly higher in northern Livland than in Estland. Differences between northern Livland and Estland also occurred in harvests and in the dynamics of grain prices.Footnote 52 Livland received a relief loan of about 1.2 million roubles from public funds, while Estland received 0.1 million. (Although the population of Livland was approximately 2.8 times larger than in Estland, the disparity is not proportional to the population.)Footnote 53
Figure 2. Increase in the average annual number of deaths in 1841–1842, 1845 and 1846 compared to the average of 1837–1839 (1837–1839 = 0).
In the regions subject to manorialism, the question of the effects of short-term economic stress on peasants’ sustenance and survival was intrinsically related to the protective role of manorial lords. There is a widespread view that the estates, for several reasons, insured their tenants against economic hardship.Footnote 54 As in Sweden, narrative accounts give many illustrations and confirmations that manors lent and sold grain (however, often at high price) to help tenants when community granaries containing spring and winter grain supplies gathered from landed peasants could not provide the needy with sufficient assistance in the form of loans.Footnote 55 Dribe, Olsson and Svensson studied the mortality response to price fluctuations in Sweden from 1749–1859 and concluded that ‘the manorial estate seems to have been able to insure its inhabitants against risks of economic stress, but the protective effect was imperfect and only short term’.Footnote 56 Analysis of relative mortality risks on six noble and one church estate in Helme and Holstre state estate (in Paistu parish, south of Viljandi parish) shows that the protective effect of the state estate was largely limited to children under five and there were no significant differences between landed and landless peasants.Footnote 57
Another possible way to measure whether the estate owners managed to protect their tenants against short-term crisis is to examine changes of tenancies. The number and names of manors experiencing different proportions of tenant turnover are presented in Table 3. If the rent contracts are available, the year 1853 is the reference year. According to court records, tenant farmers and their families still continued to be dispossessed due to crisis-time granary debts in the late 1840s and early 1850s.Footnote 58 Furthermore, in 1853, written rent contracts became mandatory on Livonian noble estates, and the reasonable estate holder most probably conducted an investigation of the tenant's capacity to meet his lease obligations and run the farm before entering into the contract. Changes in tenantship cannot, of course, be reduced to only crop failures and the resulting difficulties but were also associated with short-term leases (on noble manors), demesne expansions and tenants’ refusal to perform increased labour duties.Footnote 59
Table 3. Proportion of farms that experienced at least one change in occupancy from 1841 to 1850/1853
a State estates are in italic.
Table 3 demonstrates that the proportion of farms changing tenants varied at the level of the parish, as well as of the community (manor), but generally, the regional differences were larger than local ones. It is particularly noticeable in the less affected parishes, where the share of newcomers among the tenants was fairly similar on different estates (see Viljandi and Jämaja). In Jämaja, approximately every tenth and in Viljandi every fifth peasant left their farm. In Helme, about 40 per cent of farms changed hands. In the worst afflicted parishes of Helme and Saarde, the difference between the two extremes could be two to three times. In Martna, a more diverse picture emerges. On some estates, the occasional changes of tenancies cannot be (directly) associated with the subsistence crisis, while on other estates a quarter or more farms changed hands.
Differences in grain yields and mortality at the district levelFootnote 60 suggest that general environmental and socio-economic factors account for the prominence of regional differences. Intra-parish similarities and inter-parish differences can be partly explained by environmental conditions which did not vary much within the boundaries of a parish. Opportunities to earn supplemental income from non-farming activities were also similar within parishes. In Jämaja on the Sõrve peninsula, part of the family's livelihood could come from the sea. Forestry offered supplemental income in Saarde, where forests and bogs covered 90 per cent of the territory. These non-farming economic activities seem to have had a protective effect against mortality (Figure 2) but could not prevent frequent changes in tenancies (e.g. in Saarde).
Large farmsFootnote 61 dominated the area in Helme in which changes in tenancies were fairly frequent (Table 3).Footnote 62 In Viljandi and on the state estates in Saarde, however, large farms formed a sizable group as well; however, changes in tenancy were less frequent, and hence, the size of the farm is a rather poor explanation for variations in the permanence of farms. At the manor level in the parish of Helme, it appears that heads of large farms were relatively more successful at keeping their tenancies. In Voltveti-Kärsu in Saarde, there was no marked difference between large- and medium-sized farms.Footnote 63
The organisation of famine aid contributed to the differences between Estland and Livland. In Livland, the state government granted extensive subsistence loans to peasant communities, but under unfavourable conditions. The community was explicitly made responsible for the provision of its landless members, which, in fact, significantly raised the grain needs of the tenants given the large size of households. The farm household functioned as a welfare as well as a production unit. Widows and widowers, paupers, incapacitated veterans, soldiers’ wives, orphans and the like became farmstead members as a result of a kind of ‘embryonic social policy’.Footnote 64 In Estland, where cottagers lived on their own crofts, often on communal or manorial lands, their distress did not eventually translate into problems for the farming families, as was the case on the Livland mainland. In Estland, the number of living-in servants was smaller as well.
Unfortunately, there are very few court and other official records to give us insight into the specific factors which caused changes in occupancy in the 1840s and early 1850s. At the state estate of Pati, for example, five tenants were dispossessed in 1839/1840 for poverty (and poor health); two were evicted for granary debts and negligent farming in 1845 and 1846, and thereafter no evictions occurred until the tenancies were transformed into freeholds.Footnote 65 Their subsistence crisis of the mid-1840s led to serious arrears to the communal granary and to the leaseholder but not to many evictions.Footnote 66 At the state estate of Laiksaare, at least three tenants lost their farms because of serious debts in 1845, and in 1852, two more were evicted due to debts from the crisis time.Footnote 67 At the noble manor of Voltveti-Kärsu in the same parish, about 40 per cent of farms changed hands from 1840 to 1853 and expulsions due to excessive labour dues and/or bankruptcy continued until the mid-1850s. State peasants generally fared better than their counterparts living on noble land (Table 3). The differences, however, were not always very large, as heavy labour dues burdened both categories and the labour requirements were not in all cases substantially lower on state estates. From some manors in Helme parish, there is data on evictions due to tenants’ irrecoverable community granary debts. In 1845–1847, Lõve had four such tenants, Patküla had seven and Riidaja had none.Footnote 68 In 1849, the estate administration of Kärstna-Murikatsi reported about seven such cases. These numbers do not include all evictions due to various debts, but only evictions due to irrecoverable granary debts that were not collected from the rest of the community members through a joint surety principle.
When debts were incurred, peasants were not evicted immediately. At large farms, the landlord could add another economic unit (a co-head's family) to the peasant household to reinforce it. The tenants could find a guarantor or take a loan and thus continue at the farm. It was more complicated for debtors to maintain their farms in regions with more evident farm stratification and friction within the community (Helme and Saarde). In the case of debts due to the community granary or other communal debts, community authorities and the manorial administration (the latter oversaw the granary stocks) were sometimes very patient.Footnote 69
The following three examples from different types of manors illustrate differences in treatment by lords as well as by courts and higher officials. The case of Jaan Koik at the state estate of Holstre demonstrates how differently the Baltic governor-general treated state peasants and their peers living on noble and church estates. In 1848, Koik was evicted because of negligent farming and substantial arrears to the communal granary. The officials in charge of state domains approved his eviction, and the farm was transferred to a new tenant.Footnote 70 Koik petitioned the governor-general, who asked the parish magistrate to resolve the case, if possible, ‘in favour of the petitioner’. The farm was indeed returned to the evicted tenant, and his descendants kept it for decades. In 1847, at the church estate of Viljandi, Jüri Kiis was also evicted due to negligent farming and debts accrued during the famine years. In his appeal to the governor-general, Kiis claimed that the loans of the other tenants were deferred, while his were not.Footnote 71 His supplication bore no fruit. The lord of Uusna manor evicted Margus Vasar from his ‘father's and forefathers’ farm’, Vasara, in 1845, and transferred it to Margus’ son Jüri and then, two years later, to a stranger. Margus claimed that he was not indebted either to the lord or to the granary and had kept the farm in good order.Footnote 72 The official correspondence on this case mentions neither his arrears nor mismanagement of the farm but reveals that the lord ejected him from his holding either for converting to Russian Orthodoxy or for ‘instigating disobedience’ towards the manor's orders.Footnote 73 In essence, there was little difference between the two, as the conversion was a ‘legal’ form of protest against manorial oppression and poverty in general,Footnote 74 whereas disobedience towards the manor was not allowed. Margus's repeated appeals to the governor-general remained fruitless. The Riga archbishop also spoke up for him, by turning to the governor-general and calling his eviction ‘baseless’, but without success. After the eviction, Margus lived in his old cowshed, and in his own words, he was forced either to leave the community or to stay there as a vagrant.Footnote 75
In their complaints and petitions, evicted tenants occasionally described the fate they feared: starvation, beggary or ‘complete ruin’.Footnote 76 In the face of this threat, some became violent. In Lõve, Henrik Rebane, who was expelled from his farm by force, threatened to kill the incoming tenant.Footnote 77 Lord Anrep ejected Rebane from his holding on the grounds of ‘insufficient respect’ for the lord, based on his ‘words, tone and manners’. In Helme, an evicted farmer and his family members, the Variks, terrorised the new tenant until he left the farm.Footnote 78
Lääne county (including Martna) was already in serious difficulties in 1841–1842 and deaths were still increasing in 1846, while in other studied parishes the mortality crisis was mostly limited to 1845. In Martna, changes in tenancy were relatively less than in Helme (27 vs. 42 per cent), although parish-level mortality figures did not give a clear advantage to Martna (Figure 2). Without further research and comparative evidence from other parishes in Estland, an adequate explanation cannot be given here. It may have been that in areas with village settlements, such as Martna, the local community's safety nets were more effective in ensuring against short-term stress than in areas where single farm settlements or farm-clusters dominated the social landscape. In addition, the relative poverty and lower level of geographical mobility of peasants in EstlandFootnote 79 limited the number of competent and able candidates for headship, and thus, the current tenants may have been allowed to continue farming as there were no immediate alternatives.
The effect of famine on the permanence of farms in the study areas resulted at least partly from local lord-peasant relationships that could either provide insurance against extreme events or exacerbate the crisis. Among the manorial lords, there were both notably ‘paternalistic’ or ‘reasonable’ lords, as well as men known for their cruelty in popular memory.Footnote 80 The relative stability of tenant holdings in Jõgeveste (owned by the Barclay de Tolly family, who were known for their kindness), Riidaja, Viiratsi (owned by the town of Viljandi) and on the church estates in Jämaja, Helme, Martna and Viljandi – should be attributed mostly to the manorial lords’ ‘paternalism’ in easing the crisis in localities. Changes of tenancy cannot be explained by changes in the ownership of the manors and the demands made by new lords. In many instances, there were no changes in manor ownership.Footnote 81
The severe turnover of tenants on noble land could be explained by the manorial lord pushing the farmers to their limit and causing them to fail if they could not rent a new farm. Individual-level social and demographic data exist for Helme parish which allow us to follow the tenants who left their farms. In the manors of Leebiku, Helme and Patküla, half of such tenants did not experience downward social mobility but rather took on another tenancy. In Lõve, however, only 2 out of 14, and in Riidaja no one out of ten, were able to carry on as tenants. One tenant experienced several evictions on different manors.Footnote 82 In Saarde, population growth caused large portions of land to be brought under cultivation and new farms to be created. Nevertheless, at the noble manor of Voltveti-Kärsu, for example, half of the tenants who left their farms moved downward socially and in 20 per cent of instances, the family lost the farm upon the death of its head.
On all estates, either locals or residents of nearby estates dominated among the new tenants. Only a few came from more distant places and they usually originated from more developed regions like Mulgimaa and the vicinity of Viljandi. Economic downturn increased opportunities for upward mobility for individuals from the lower social groups, especially able-bodied farmhands, although their share varied. Frequent change of tenancy and the rise from farmhand to tenant farmer was partly facilitated by the fact that the ‘iron inventory’ of the farm stayed with the farm and the outgoing tenant had to leave the farm in good condition. A sample of five noble manors illustrates this point. In Jäärja a half, in Ala-Aitsra and Kärstna-Murikatsi a third, in Leebiku-Vanamõisa a fifth and in Päri very few, new tenants had formerly been servants.Footnote 83
7. Introduction of money rents and farm proprietorship
The change to money rents and the purchase of farms advanced across Estonia at very different rates. In the area of investigation, the transition to money rent was completed on state estates in 1854, while most farms on noble lands transitioned in the mid-1860s.Footnote 84 Even after the ruling banning the corvée in 1868, the use of partial corvée in Estland and Ösel was allowed. During the transition, the timing of this new mode of rent affected the change of tenancy only if it coincided with crop failure or famine years. Otherwise, the first money rent agreements were mostly concluded with the current tenants. In the parish of Saarde, however, there were many newcomers among the tenants.
The stability of tenant farmer households on state estates in northern Livland and Ösel is not surprising, as the rent was affordable and the state government could not manage to evict debtors.Footnote 85 On noble estates, as opposed to state estates, changes of tenancy also continued during the time of money rent. Most common were evictions due to poor farming or moral deviance (e.g. poor cultivation of the fields, bad condition of the dwellings, debts, opposition to the increase in rental prices, non-compliance with the provisions of the rental contract (e.g. sub-renting farmland or selling hay without obtaining manorial consentFootnote 86), ‘various complaints’,Footnote 87 wood and other thefts,Footnote 88 excessive drinking,Footnote 89 ‘moral corruption’, negligence and laziness,Footnote 90 and contrarinessFootnote 91). Court records reveal that estate administrations kept a careful watch over the tenants and were well informed about how much rye or flax each tenant had sown, how many trees he cut, how much cattle food he sold and so on. The lord of Jäärja manor used the sale of a single haystack without his permission as an excuse for evicting an undesirable and troublesome tenant.Footnote 92 An accusation of ‘damaging the farm’ could arise from the tenant's attempt to reorganise production in response to market demands (extensive cultivation of the cash crop, flax; clearing land and reducing pasturage; and raising more cattle than the rent contract allowed), which led to evictions at different manors.Footnote 93 The extent of manorial control in northern Livland and Estland seems to significantly differ from that revealed, for instance, about early modern Bohemia,Footnote 94 but for more useful comparisons, we need empirical findings on Bohemia from later periods.
On noble lands in Ösel and large parts of Estland, the transition to money rent coincided with the famine of 1868–1869, and any possible effects of these two events cannot be separated from each other. In Martna, however, neither of these resulted in a severe turnover of tenants. In 1868, in the parish of Jämaja, one tenant in Kargi and four tenants in Kaunispe did not renew a rental contract; on the church estate, all farmers stayed on as tenants. Perhaps the tenants who stayed at their farms started to pay rent in cash only on paper.Footnote 95
Land purchases proceeded more slowly in Estland than in Livland and Kurland. By the early 1870s, 20 per cent of farms on noble estates in Kurland and about 25 per cent in Livland had been bought outright by peasants.Footnote 96 In Estland, the respective figure was 111 holdings (less than one per cent).Footnote 97
Table 4 displays the frequency of changes in farm occupancy related to the buying of land outright in the northern Livland mainland. In most cases, a quarter or more of farmers lost their farms, despite having the first option to purchase. The number of changes of occupancy in northern Livland contrasts with respective figures in Kurland, where nearly 80 per cent of the buyers of farmsteads between 1865 and 1885 were the current tenants of those farms.Footnote 98 Here, the most intensive period of land purchases was at a time of high grain prices in the late 1870s and early 1880s, during which time the grain export from Kurland to Riga also grew.Footnote 99
Table 4. Proportion of farms that changed occupants due to outright purchase of land, 1850–1889
a Data on Suure-Rõude, Suure-Lähtru and Laiküla where a considerable number of farms were bought outright before 1889.
The pace of purchasing farms on noble land was determined largely by regional factors. Whether former tenants could buy their farms in perpetuity depended primarily on local circumstances. The time of sale, the landlord's attitude and the price he wanted for the farm, competition from neighbours and migrants during the sale, the share of previously in-migrated farmers, purchases of farms and demesne expansions in the surrounding regions, all played a role. For example, the lord of Jõgeveste manor, Magnus Barclay de Tolly, sold the farms to former tenants at a very favourable price, only about half the average in that region. Karl von Bruiningk, by contrast, tried to sell a large portion of his manors’ farmlands all at once in a time when current tenants did not have the pre-emptive right to land purchase, tenants were low on resources and the Livonian noble credit bank was not yet offering mortgage loans to peasants. As a result, in Ala-Aitsra approximately half and in Leebiku-Vanamõisa over 80 per cent of farms changed hands. In many places, tenants asked the landlord to delay the sale, so that the farms could be improved or so that the tenants could at least obtain a loan. There were manorial lords who agreed to wait for ten or more years, while others refused to wait even one year.Footnote 100 In Mulgimaa, the allocation of one-fifth of the farmland (so-called Quota land) for the use of the manorial lords as well as the sale of farmsteads triggered the outmigration of evicted tenants and the purchasers’ family members (brothers, younger sons and others) who sought a new farm outside Mulgimaa.Footnote 101 Christine Marie von Frey, the lord of Tali manor in Saarde parish, wishing to sell farmland to local peasants, was rather patient with tenants who at first refused to agree to purchase conditions. At the same time, interested buyers from Mulgimaa persistently urged her to sell the farms to them and offered a higher price than she asked from locals. After she had signed purchase contracts with some newcomers, locals, out of fear for losing their farms, started to buy the farmsteads.Footnote 102 Competition by migrants varied in intensity from manor to manor.Footnote 103
In Helme parish, half of the former tenants whose farms were sold but who could not or did not buy them outright themselves either fell into the ranks of landless and semi-landless (cottagers) or migrated out.Footnote 104 In Päri, about 60 per cent of the former tenants continued as farmers in the same parish, and about 40 per cent moved downwards socially. In Vana- and Uue-Pärsti, one-third of former tenants managed to retain their status as farmers by acquiring a new farm. The parishes of Helme (and Mulgimaa broadly) and Viljandi were in a much better position than most other parishes in Estonia as many tenants whose farms were sold to others migrated out and managed to rent or buy a farm elsewhere, which tenants in less fertile and well-off regions could hardly afford.Footnote 105
The sale of farmland to strangers sparked petitions and protests in the first half of the 1860s. Earlier research has suggested that the ‘fight for one's own farm’ gained momentum with the sale of farmland in perpetuity and tenants clung to their farms, referring to ‘their only property’ and their ‘forefathers' heritage’.Footnote 106 Attachment to one's own farm is also reflected by the fact that an evicted tenant refused to accept another farm offered by the landlord.Footnote 107 In Kärstna-Murikatsi, 14 farms were sold to strangers and at least 10 dispossessed tenants either turned to the court or the governor-general. Six of them claimed that they had improved their farms ‘in the hope that they and their children retain them’.Footnote 108
In the petitions to higher authorities in St. Petersburg and Riga, tenants complained of high prices demanded by manorial lords and their lack of money to complete the purchase. In 1864, the representatives of 16 communities (including 6 from our study area) claimed in their petition to the Emperor that ‘evictions from their fathers' farms’ led tenant families to ‘great poverty and misery’.Footnote 109 Regulation of land prices demanded in petitions was not achieved,Footnote 110 and purchases remained within free-market conditions. In the mid-1860s, resistance to farm-sale-related evictions was strongest in the parish of Viljandi. For example, tenant farmers from Puiatu, Suure-Kõpu, Karula and Viljandi collectively and repeatedly protested against the sale of farms to strangers instead of current tenants.Footnote 111 In Päri, evicted tenants continued to occupy their farms for months after the formal eviction.Footnote 112 One such tenant stayed on his farm for seven years until the court ruled again that he must leave.Footnote 113 A farm could pass from hand to hand if a current tenant, who at first could not find the resources, managed to mobilise them after his farm had been transferred to another buyer, and the landlord agreed to return the farm to the tenant. Then, the disappointed buyer had to seek justice in court.Footnote 114
In the 1860s, as also happened earlier,Footnote 115 evicted tenants refused to leave the farm and communal authorities, estate administrators or members of the parish court expelled them by force.Footnote 116 The executors did not refrain from using physical violence, and in one such incident, they even shoved a woman who was nine months pregnant.Footnote 117 Family members of the tenants occasionally physically resisted the eviction, and on the manor of Viljandi, one such clash resulted in eight months’ imprisonment for two women. One of them had poured boiling water on the soldiers who had climbed over the fence.Footnote 118 If the police could not do the job themselves, they asked soldiers for help in evicting farming families.Footnote 119 Cossacks were also stationed on some manors in order to ‘calm’ the situation.Footnote 120 Conciliation, however, was preferred to military repression.
In all three studied parishes of northern Livland, the new owners who purchased the farms were split between locals and migrants. Most of these migrants were either from other manors in the same parish or from Mulgimaa, while the rest were almost always either from nearby Viljandi or Saarde parish bordering on Mulgimaa. In the parish of Helme, mostly well-off and astute men from elsewhere in Mulgimaa managed to establish themselves as owners in addition to the locals. On the noble estates in Jämaja and also mostly in Martna, farms were bought outright at the turn of the century, and subsequently, farms also frequently changed hands. On the noble estates of Ösel, little more than half of the former tenants managed to purchase the farms.Footnote 121
The sample of eight estates (Jäärja, Päri, Ala-Aitsra, Helme, Kärstna-Murikatsi, Leebiku-Vanamõisa, Lõve and Riidaja) shows that the opportunities for a servant to advance to peasant status by buying land were very limited. It happened only in few cases in Jäärja and Lõve. The social status of buyers in the years prior to the purchase of the farm (and not at the time of birth) is used as an indication of social status. Many buyers were born into servant families, but they had changed social status before their farm purchase. Their upward mobility into the ranks of farmers, inn-keepers, storekeepers, teachers or bailiffs did not result from buying land.
8. Conclusion
The data presented in the article suggest that farmer families in Livland and Estland had difficulties with keeping the farm within the family. Apparently, this was not the case in Kurland. There, the subsistence crisis in the 1840s had less of an impact, and one-third of the peasants lived on state lands where tenure rights were more secure. Even on noble lands, the majority of tenants bought their farms in perpetuity. In Estland and Livland, there were many factors influencing the permanence of farms in various ways at different times. Whether a farm was retained or not depended on a combination of institutional, socio-economic, demographic and ecological circumstances and only by combining analyses of their interactions can we get a better understanding of the inner workings of the manorial economy and the farming households.
In the five parishes under study, in most if not all cases, the changes of farm occupation were quite frequent on noble manors. Farms became prominently more permanent on state estates than on nearby noble estates in times of transition to money rent and land purchases. Thus, institutional setting – tenurial security and strong hereditary rights – significantly affected the permanence of farm tenantship or ownership. A comparison of the crisis and reform periods reveals that legal protection on state estates could insure the stability of farms in times of scarcity much less than in the time of agricultural transformation. In the 1840s and early 1850s, farms also changed hands due to arrears, negligent farming and impoverishment on state estates, but in the time of agricultural transformation, changes of occupancy happened very seldom. The relatively low turnover among tenants in Martna in Estland and especially Jämaja in Ösel, where tenants lived on small farms and cottagers in their separate households, suggests that in times of short-term crisis, settlement patterns and household size also mattered. Village systems promoted mutual insurance networks within a community. In the time of buying farms outright, the situation changed substantially. On noble manors, tenants needed substantial resources of their own to retain the farms, making community safety networks less important. Paternalistic bonds between manorial lords and tenant farmers – or the ‘reasonable’ actions of manorial lords – could positively affect the permanence of farms, but this potential was actually little used, as the exceptional cases of Viiratsi, Jõgeveste and church estates seem to suggest. Estonian data casts doubt on the widespread idea that manors efficiently insured tenants against economic hardship in preindustrial times.
Remarkable local exceptions of very high permanency rates allow the conclusion to be drawn that the main reason for leaving was manorial pressure. Larger differences within the parish emerged after the disappearance of corvée. The impact of socio-economic factors, such as farming and market conditions, varied in time and place and as relationships between manorial lords and tenant farmers changed. The situation was easier for farmers in some of the more well-off areas, such as Viljandi parish (but the example of Helme and Saarde shows that this was not the case generally in such regions), and on the periphery where the degree of commercial orientation of production was minimal and there was little outside pressure from better-off and ambitious newcomers. On the Jämaja church estate, where even peasants practised a two-field system, tenancies were the most permanent in the study area. In all parishes, permanency was related positively to secure tenure. In Ösel, it was also positively related to unsuitable environmental conditions and limited market opportunities before the transition to outright purchase of farms. The outright purchase of land required large sums of capital; thus, good market opportunities could favour permanence of farms (e.g. in Kurland).
The regional pattern of change of occupancy cannot be explained with ‘distinct lifestyles’, as Christiansen has suggested in his study of Denmark. Given the evidence it is impossible in the case of Estonia to reduce a widespread phenomenon such as change of occupancy to people's individual qualities and qualifications. It would be difficult to claim that the people at places with continuous farm holdings were more capable, careful, austere and responsible or that they were protected from weather fluctuations, illnesses and accidents. In this historical context of severe restraints, seeing peasants solely as pursuing ‘strategies’ or as rational economic actors is misleading. The existing data lend strong support for the idea that in favourable conditions (where either the manorial pressure or competition for farms were lower), tenants kept the farms in the family, and thus in the short term, the frequent comings and goings of families on other manors were neither ‘naturally’ inevitable nor mostly demographically determined.
Fairly frequent changes of occupancy on noble estates, in contrast to state estates, reaffirms the traditional view of the exploitative character of landlord-peasant relationships and the location of Estonia among areas of highly insecure tenure. The relative power of manorial lords vis-à-vis tenant farmers in terms of land tenure was strong not only in theory but also in practice. This finding is in accordance with the studies on eastern Denmark, Schleswig-Holstein and Scania, which note the instability of tenant holdings in the face of manorial pressure.Footnote 122
Secure tenure and control over farm succession were vital economic incentives for pre-industrial tenant farmers, and the noble politicians were concerned that if the farmers were afraid of losing their land they would not make long-term investments in farm production and neglect the farm. The decline of mutual economic dependence between manors and farms, so that only a cash nexus in the form of rent or mortgages between the estate owner and the farmer peasant remained, changed the virtue of stability for a short while. The lords of noble manors tried to rent and sell the farms to the best-suited candidates and restrict the number of under-achieving farmers. In this short transition period, stability could no longer be viewed as a value on its own because the fluidity of land ownership generated prerequisites for production resources to go to the most well-off purchasers. In socio-economically more developed parishes like Helme, the earlier start of the agricultural transformation (improved tools, crop rotation systems, and commercialisation of farm production) was facilitated by transferring assets into the hands of the most forward-looking peasants, who concentrated on working with their resources to improve the farm's economic viability. In the time of outright purchase of farmsteads, changes of occupancy could strengthen farmers’ competitive strength. However, in the long term, it was the secure ownership after tenancies had been transformed into freeholds that provided vital economic incentives for the farmers to further develop the farm economy.
Recent literature on pre-industrial rural societies and peasantries in Europe has jettisoned the idea of peasant immobility and my findings support this conclusion. At the same time, however, the case of Estonia lends support for the conventional view that if circumstances beyond the control of farmers allowed, they tended to live within the borders of the same community,Footnote 123 keep the farm within the family and be succeeded by their offspring. Taken together, the data here suggest that the relative weakness of tenant farmers vis-à-vis landlords increased the geographical as well as social mobility of the landholding groups.
Acknowledgement
The author would like to thank the editors for their useful comments. This work was supported by the Estonian Research Council grant PUT-1517 and Herder Institute Fellowship.
