The Weights and New Products

The inclusion of rents and the price of bread is a major step forward in comparison with the first-generation CPIs. The selection of different weighting schemes is another potential area of improvement.

These four expenditure patterns were applied to the price series, which eventually were converted into indices 1450—74=100. To find out how important the use of different weighting schemes is, I calculated four CPIs on this basis (1450-74=100) (see Table 7.4). The index of wages was used as the price index of consumer services in the elite CPI, which represents the ‘consumption' of the labour of servants.

The similarities between the different series are perhaps more striking than the differences. The two indices based on the estimated expenditure patterns of labourers in Holland show a remarkable degree of similarity. This is an important result as it shows that, in spite of the fact that the labourers tried to adapt to changes

Table 7.3 Stylized expenditure patterns

‘ In the elite index this was estimated as sugar 5% and pepper 5%.

^b Estimated by the price of paper.

Based on: nineteenth-century budgets: van Zanden and van Riel (2000: 409), other budgets: de Vries and van der Woude (1997: 708—9), Phelps Brown and Hopkins (1981).

in relative prices by adjusting their demand patterns (i.e. by consuming less expensive bread, meat, and butter, and more cheap textiles), they were unable to do much about the enormous increase in the general price level. An important reason for this was that they were ‘forced’ to spend more on housing, due to the fact that rents were going up relative to all other prices.

The absence of this rent squeeze, which has been ignored in much of the literature on the price revolution of the early modern period, largely explains the difference between the England price index (i.e. the price index using the PBH weights) and the Holland indices, although the substitution of bread prices for rye also played a role. The elite CPI also shows a much less rapid increase, which is mainly due to three developments: (1) bread had a much smaller weight in the basket of the elite; (2) the elite profited from the relative decline of the price of consumer services (wages); and (3) they also profited from the decline of the prices of industrial and colonial products, which they consumed in much larger quantities than the

Table 7.4 CPIs using four different weighting schemes, and the final index, 1450/74—100

Sources. Tables 7.2, 7.3, and Appendix.

labourers. This experiment suggests that in the early modern period, relative prices moved in general in a direction which was unfavourable for the lower classes (i.e. sharp increases in the prices of bread, beer, and of rents) and favourable for the elite (i.e. a relative decline of wages—and therefore of household services—and of industrial prices).

Changes in the patterns of consumption of the labourers did not affect their CPI very much, however. The strong decline in meat consumption and the increased demand for industrial products that can be observed in early modern Europe were rather weak attempts to stem the tide of the almost continuous rise of prices (only after 1675 did inflation come to a halt). Similar results were obtained by Phelps Brown and Hopkins (1981: 89), who in one of their papers also referred to an experiment with different weights for their CPI, but which did not have a big impact on its long­term development. This is a striking result, because one would expect a rather different outcome, that is, that changes in relative prices and demand patterns combined would have moderated the price increases to some extent. Labourers were, however, ‘trapped’ because they really needed basic foodstuffs and shelter and could not switch out of them to an extent which would seriously have changed their ‘terms of trade’ (i.e. the relationship between nominal wages and the CPI).

There may have been one exception to this. In the literature on changes in patterns of food consumption in this period much attention has been paid to the rise of the potato, a crop which replaced bread and vegetables (peas and beans) on the tables of labourers in the second half of the eighteenth century. This occurred in a period of renewed inflation when Dutch nominal wages remained more or less the same (and

therefore real wages seem to have declined substantially). Noordegraaf and others have suggested that labourers may have stabilized their real incomes by switching to potatoes, which provided many more calories per guilder.

There are different ways to change the CPI in order to take this process into account. The standard procedure is to create new indices on the basis of the new expenditure patterns which come into existence, in this case, as a result of the introduction of the potato; on the basis of the qualitative literature on this topic and the nineteenth-century budgets it can be estimated that the share of potatoes in the budget increased from 3% in the 1770s to 6% in the 1780s and 9% in the 1790s.

A ‘quasi-hedonic’ price index can also be estimated: assuming that the aim of consumption (of bread and potatoes) is the intake of calories, one can calculate the relative prices of the calories contained in these products, and introduce the price of potatoes accordingly. The welfare effect of getting more calories per guilder from potatoes is then taken into account (the weights are again identical to the ones used in the first approach; 3% for the 1770s etc.). Figure 7.1 gives an indication of the kilocalories that one guilder represented, which clearly shows the big difference in price per calorie between potatoes on the one hand and bread and peas on the other hand.

Both adaptations of the CPI show a more moderate rate of inflation after 1763 when the series of potato prices begins. If we accordingly recalculate the CPI based on the eighteenth-century weights, and set 1763=100 for convenience, the index for 1790—9 is 122.9 for the unchanged CPI, 119.5 for the CPI with a conventional introduction of the potato, and 115.9 for the recalculated CPI with a ‘hedonic’ introduction of the new crop. It appears that the potato did make a difference, but it only lowered the rate of inflation after 1763 by about a third.

Figure 7.1 The price of 1,000 kcal in guilders (1783—1800)

Figure 7.2 The CPI for the western part of the Netherlands, 1450—1800 (1450/74—100; polynomial trend added)

The final CPI was calculated on the basis of the fifteenth-century weights for the period 1450—1656 (1450-74—100) and on the basis of the eighteenth-century weights for the period 1656-1800 (1720-44—100).¹⁰ The final index (recalculated 1450/74—100) is shown in Figure 7.2.

5.