Janken Myrdal - The Agrarian History of Sweden: From 4000 BC to AD 2000

For the earliest period the percentages in refer to the number of farms, but thereafter they refer to mantal or hemman ; two terms with partly overlapping meanings, where hemman originally meant physical farms and cadastral units while mantal measured the taxable capacity of the farms.

The nobilitys land was tax-exempt ( frlsejord ), whereas freehold peasants land was taxed ( skattejord ). By the mid seventeenth century, however, about a third of the land controlled by the nobility was comprised of freehold farms where the nobility had acquired from the Crown the right to tax the freehold peasants who farmed thembut not full ownership of the farms in question. Of the remaining 35 per cent of the land, roughly half was in the hands of the Crown and the other half in the hands of freehold peasants.

The first historian to summarize the landownership figures was Eli Heckscher. Subsequently his results have been revised several times. For the eighteenth and nineteenth century his figures are misleading: he calculated by cadastral category (tax-exempt land, Crown land, and taxed land) and disregarded the fact that an increasing share of the tax-exempt land, originally owned by the nobility, passed into the hands of the bourgeoisie and non-noble persons of rank, and finally even to freehold peasants.

From the mid nineteenth century Swedish landownership was not normally related to social class, while the cadastral categories of agricultural land had by then lost much of their social connotations.

References:

Carlsson 1956, 192; Carlsson 1973, 170; Ferm 1990, 387; Heckscher 1935, app. 4; L.-O. Larsson 1985; Myrdal 1988; Nilsson 1964, 88; Rydeberg 1985, 162.

Percentage landownership by social category, 15201845.

Source: Myrdal 1988.

During the eighteenth and nineteenth centuries the social structure of the countryside underwent radical changes. Notably, there was an increase in the number of households with no access to land or (insecure) access to very marginal amounts of land. Statare (see pp. 141 and 175) exemplified the first category (farm-hands living in the peasant-farmers or farmers households are not independentlycounted). This development is shown in shows, although their decline started later than other categories of unlanded labour. These other categoriescrofters, day labourers, farm-hands living in farmers households, etc.outnumbered the statare , especially before 1870.

Number of peasant-farmer, crofter, and cottager households, 17511910, and statare households 18301943.

Sources: Wohlin 1909, sammandragstabell C, D, E, rtabeller for peasants, crofters, and cottagers; Bagge, Lundberg & Svenilsson 1933, 1945 for statare 18701920; Furuland 1962, 175 for statare 1930; M. Olsson 2008, 71 for statare 1938 and 1943.

Number of agricultural holdings by size of arable area, 18852004.

Sources: Jordbruksverket 2005a, ; for number of crofts 18901900, Wohlin 1909, raw data tables; for 1910, SOS 1910 Folkrkningen ; from 19191943 crofts were included among other units and not registered separately.

Table 2.3 Percentage arable acreage by size of holding in hectares, 19191999.

Sources: SCB 1920; SOS 192737, Jordbruksrkningen ; Jordbruksverket, Jordbruksstatistisk rsbok 19702000.

Several historical GDP series exist. Hitherto, agricultures contribution to GDP has been underestimated for the early nineteenth century, for although the official underestimation of acreage (discussed in greater detail in section 4 below) was compensated for, the proportion of harvests used for forage and thus as input in animal production was grossly overestimated, leading to a decreased market value of total production. The series presented here are duly revised, taking into account provisional recalculations of grain harvests made by Carl-Johan Gadd (Gadd 2008).

Data on agricultural employment is notoriously problematic, as many groups worked part-time there are immense problems with arriving at a definition of the agricultural work-force. For example, many farmers and farm workers worked in forestry or had industrial occupations on the side, and the farm work performed by farmers wives and family members is an uncertain point. Above all, the distinction between household work on farms and agricultural work is very unclear. The figure for agricultures share of total employment given here is therefore very approximate.

Agricultures percentage share of GDP at fixed prices, 18012000, and of employment, 18512000 (ten-year averages).

Sources: for GDP the data was generously provided by Rodney Edvinsson of Stockholm University; for employment, LU MADD 18502000.

The first attempts to construct national agricultural statistics were made in the early nineteenth century when parish priests were requested to collect not only population data, but also figures for arable acreage, amount of barrels sown, yields, and number of farm animals. It has been shownusing cadastral maps and quantities of seed deducted from probate inventoriesthat the clergys data on acreage, seed, and total harvests involved gross underestimations. However, their data on the proportions of different grains, seedyield ratios, and number of farm animals are believed to be reasonably realistic. Around 1820 the responsibility for agricultural statistics was taken over by the county administrations and the results were printed in the county governors quinquennial reports. In 1865 the responsibility was shifted again, this time to the county agricultural societies. The quality of figures improved gradually, but the figures for arable acreage (and implicitly for totalharvests) remained the weak point until the early twentieth century. While, for example, the figure for the acreage given by the clergy in around 1805 needs to be increased by about 100 per cent, official statistics for the arable acreage of the 1860s need only be increased by about 30 per cent. As data collection methods improved, official figures in time came closer to reality, giving the impression that there was a strong increase in acreage even in the late nineteenth century (see Svensson 1965). By the early twentieth century, official data appears reasonably correct. These observations have informed both . in the previous section.