A New Approach to Standard of Living
We measure the standard of living by the ability to overcome short-term economic stress. If one cannot fulfil one's long-term plans—to survive, to marry, and to have children—in the face of acute short-term changes in the environment, one can be said to have a rather low standard of living.
Thus, sensitivity to short-term economic stress reveals a low level of standard of living. By short-term economic stress, we mean variations in income or the cost of living, particularly in food prices, from one year to the next or even within short time spans. Variations in food prices were often substantial and possibilities to compensate by increasing income were few. Temperature data could also be used as an alternative indicator of short-term economic stress. Because poorer households often could not afford to heat their houses in winter, the members were consequently more susceptible to disease, and were in frailer health than members of wealthier households. But short-term economic stress could stem from other sources as well, such as highly virulent epidemics, changes in taxation, wars, etc. There is no doubt however, that the changes in harvests and food prices were of great concern for a majority of the population up to the twentieth century, which is the reason for choosing food prices as the indicator of short-term economic stress.4 Most people spent a major share of their income on food, and bread dominated the diet.5To justify the choice of grain price as an indicator of short-term economic stress, it is necessary to establish a link between food prices and consumption. Generally, it is believed that the change in production is smaller than in price (Wrigley 1987:93; Livi-Bacci 1991:61; see also Bengtsson 2004). Even though large price changes—30% or more from one year to the next—were common, changes in production were somewhat smaller.
How much then did the calorie intake change with such changes in grain production? While we do not have any calorie calculations for southern Sweden, we have it for nearby Copenhagen. The yearly decline in calorie intake per head in Copenhagen could be as much as 14% during the worstyears of the eighteenth century (Thestrup 1971: 258—9). This figure is based on the assumption that the quantity and quality of other food items, like pork and beef, were not influenced by the bad harvest, which is optimistic. Not all people had to lower their consumption in bad years, however some inhabitants had to lower it more. A qualified guess is that a decline in the calorie intake for the poorer parts of the population by about 10 to 20% was a likely outcome of quite normal price increases during the pre-industrial period. In years of very high prices, the effects are bound to have been much bigger for these groups while other groups certainly benefited from high prices.
Since people in the past were well aware of the short-term variation in harvests and demand for labour, they planned for it. The most obvious measure taken would be to store food, but storage was very costly in the past, since a large part of the stored grain was lost while in storage. As much as 20% per year might be lost when storing grain; for potatoes the figure would be even bigger. Moreover, the nutritional content of the stored food was degraded. Consequently, the quantities of food stored in the past were only very limited (e.g Persson 1999: ch. 3). Saving money to buy food could be costly as well, since prices went up in years of bad harvests. Diversification of production was therefore necessary. When harvests failed, efforts were made to find additional sources of consumption in other vegetables and meat. It is, however, unlikely that it was possible to fully compensate the gap in consumption this way. Another potential way of smoothing per capita consumption was to send some family members away to seek food and/or employment elsewhere.
That way what food there was lasted longer for the remaining family members, and hopefully the migrating family member could bring something back as well.If these measures failed, the family could borrow food or money to maintain consumption. Kin, neighbours, employers, and churches gave loans for consumption. Borrowing from the church or a bank depended, however, to a large extent on access to land or other assets that could be used as collateral, or security, for loans (Svensson 2001). Relatives and employers were more likely sources of help. Employers had a legal responsibility to take care of their workers, as had parents to take care of their children and vice versa when the parents became old.
If all the measures taken by employers, family, and kin failed, the local community or state might provide help. The poor law system, which was a local responsibility, became more regulated all over Sweden during the 1830s and the 1840s (Skoglund 1992). However, it was intended for only a few categories of the needy, amounting to a very small percentage of the population: the utterly destitute, often elderly and/or handicapped persons with no resources at all.6 Other social welfare measures, such as extra social spending on roads and other public works in years of high unemployment, were modest. The spending on creating new jobs averaged around 1% of the annual governmental expenditures with a peak of 2.6% in 1846 (Olofsson 1996: 121).
Naturally, measures to smooth consumption are likely to have differed between different groups in society. Therefore, one important question is how different
individual and household characteristics, such as land-ownership, social status, position in the household, etc., influenced the demographic response to economic and demographic stress. For example, households that farmed their own land and employed servants or labourers, were probably in a better position to avoid the most serious effects of economic stress, while those totally lacking resources had much smaller opportunities to undertake such measures, which made them highly vulnerable to stress.
The idea of analysing the demographic response to short-term economic stress in different social groups by combining longitudinal micro-demographic data with macro-economic data was first developed by Bengtsson (1989, 1993) and has been refined within the EAP project, in particular when it comes to the importance of household characteristics and transfers within the household in conditioning the response to stress (Bengtsson 2004). Although this standard of living concept can also be used in modern welfare societies (after suitable modification), it has been developed for situations where data are sparse and where living standards are low. The ideas originate from analyses of the effects of economic stress, measured by food price variation, on mortality, fertility, and nuptiality carried out at macro level and bear some similarities to Amartya Sen's concepts ‘functionings' and ‘capabilities', as shown in the introduction of this volume.
4.