A city consumes products from its tributary hinterland and from more distant sources in exchange for goods and services. Madrid increasingly consumed a distinctive mix of commodities and provided few commercial or manufacturing services in return. Unequal income distribution heightened the effect of urban elasticities of demand and limited the incentives the city offered to its hinterland, despite the growth of population. Such growth aggravated recurrent subsistence crises and shortages and raised supply costs and food prices, forcing the poor to forgo other regional products for basic foodstuffs. For different reasons, the small elite maintained a similarly inelastic demand for imports. Demand for middling-quality farm products and manufactures from the hinterland was thus vulnerable to economic change and was correspondingly unstable. The urban market also changed as the fortunes of the monarchy altered the city's economic base. Finally, the geographic scope of the city-oriented market depended on relative transport costs and on the willingness of the authorities to coerce commodity movements.

Generally speaking, the commodities that are best documented are the basic staples drawn from the agricultural hinterland of Old and New Castile, some of which can be traced from the late sixteenth century. Not until the eighteenth century do we have data on the luxury goods and colonial products consumed by the city's elite market.
 
[109] I. Consumption of Basic Staples, Sixteenth Through Nineteenth Centuries
 
A. Wheat

Wheat was not only the largest item in the budget of three-fourths of Madrid's population, but represented 30% of the value of imports from the city's hinterland and over 10% of all urban imports. Consequently, its price and availability played a major role in shaping demand for other commodities. ⁽¹⁾ Unfortunately, there are few concrete figures for annual consumption before the nineteenth century. Those that we do have are given in Table 6.1, with indications of the relative quality of the harvests. ⁽²⁾ The growth of demand before 1630 is clear, although the decline of the mid-seventeenth century is less well documented. The eighteenth-century expansion is well defined, as is the contraction of Madrid in the early nineteenth century.

We get a fuller picture of the actual consumption levels from Figure 6.1, which extrapolates from reported crisis consumption to suggest the scope of the market under normal supply conditions. Demand for wheat and population were closely connected and reinforced each other. Consequently, the population estimates from Chapter 2 have been superimposed on Figure 6.1. The range of variation of total consumption in response to harvest quality is based on eighteenth- and nineteenth-century sources. In 1767 officials reported that a 20% decline below normal supply was enough to cause starvation, while demand in a bountiful year might exceed normal levels by 30%. ⁽³⁾ Similarly, annual totals in the 1820's and 1840's show bad years that were 23% and 30% below the prevailing average. ⁽⁴⁾ Thus crisis-level consumption was about 75% of normal, and the band reflecting normal to minimal wheat consumption on Figure 6.1 is based on that assumption.

With certain exceptions, fluctuations in wheat consumption coincide with long term trends in wheat prices, an important link between urban and regional economies. In the late sixteenth century, nine-year wheat prices rose steadily to a peak in 1595-1600, eased off, then hit a higher peak 1606-1610. Thereafter, despite rapid growth in Madrid's wheat consumption, the trends diverged for the next fifteen years and wheat prices drifted downward until the bad weather of 1626-31 produced the highest averages in the first half of the century. The price then drifted downward until 1642 as urban growth slackened, followed by a rise to political and fiscal conditions in the 1640s. The latter period experienced the highest single year prices before 1650, but the sharpest short-run fluctuations came in 1598, 1607-08, and 1626-31. ⁽¹⁰⁾ The early seventeenth century saw a break in the long sixteenth century trend toward a tighter grain market in New Castile that requires examination. As we will see in chapter 10, that break was connected with the collapse of Toledo as a competing market. 

[112] Population and wheat consumption stagnated in the 1630's, and the 75 years after 1650 saw a downward trend in the price of wheat despite economic and political instability. Falling to the level of 1601-20 after 1650, prices fluctuated violently around that trend during the crises of 1664,1668, 1677-78, and 1683-84, and the recurrent monetary manipulations. In the background, however, the basic level of urban demand declined with population. ⁽¹¹⁾ The deflation of 1680 reduced nominal prices sharply and, except for brief shortages in the 1690's, prices, population, and consumption changed little for a quarter-century. Poor harvests and the Succession War forced prices up for a few years, but thereafter wheat prices reached remarkably low levels in the 1720's.

In the second quarter of the new century Madrid began to expand slowly, and by 1740 an upward trend in wheat prices had developed that continued throughout the eighteenth century. ⁽¹²⁾ With accelerating population growth and wheat consumption, wheat prices exhibited a series of plateaus separated by increases in the basic price level in 1730-36, 1762-66, 1781-83, and 1799-1803. As in the early seventeenth century, the averages obscure year-to-year fluctuations, and there were extremely high prices in 1753, 1766, 1780, 1794-95, and 1803-04. ⁽¹³⁾ The tendency for fluctuations to become more extreme is apparent, suggesting that once more consumption was reaching the limits of available supplies. ⁽¹⁴⁾
 
B. Wine
 
Because it was regularly taxed, evidence for trends and volume in wine consumption is much fuller than for wheat. Figures for the volume of wine entering the city were found for the 1630's, 1698-99, 1730-40, 1757, 1770-80, and 1789, and for 17 years between 1824 and 1847. Other sixteenth- and seventeenth-century figures are based on the sisa del vino, imposed in 1582 at [113] the rate of 16 maravedises per arroba. ⁽¹⁵⁾ Comparable estimates for the eighteenth century are derived from the revenues of the primer cuartillo en arroba de vino. ⁽¹⁶⁾

Figure 6.2 shows rapid growth in wine consumption before 1630, a gradual decline in the next decade, and an abrupt drop by midcentury. Consumption leveled off between 1655 and the crises of 1677-85, then dropped even further. After 1730 it fluctuated around a gradual upward trend until 1790, when consumption stagnated. After the French occupation (1808-13) it rose for a decade, then declined in 1825-40, before it began a more sustained expansion.

While wine consumption was affected by population, the correlation was weak. It rose almost as quickly as population until 1630, but then declined more sharply than population. ⁽¹⁷⁾ After 1730 consumption responded more weakly to population change than it had a century before. In the 1730's Madrid consumed 475,000 arrobas per year, and about 10% less in the 1750's. From that point consumption climbed to 550,000 arrobas in 1789, then dropped back to 500,000 by 1807--only 10% greater than in the 1730's, even though the population was 40% larger. After a high in 1814-20, the figures ran around 550,000 arrobas in the 1820's, fell dramatically during the [114] Carlist War, then rose steadily during the next decade. ⁽¹⁸⁾ The long-term trend suggests that two factors were affecting total consumption: the level of population and a decline in per-capita use of wine.
 
C. Olive Oil
 
Madrid's consumption of olive oil also reflects population change, but suggests an increase in individual usage over 250 years. The sixteenth and seventeenth century figures in Figure 6.3 are based on the sisa del aceite established in 1582 at the rate of 100 marevedises per arroba. Later eighteenth-century estimates are calculated from oil-tax revenues and the ratio between revenue and known consumption.

Most olive oil came from Andalucía, and was produced under regional conditions different from those of Castile. Thus short-term changes in consumption do not always coincide with those of wine, as when plague struck Andalucía in 1590, 1626-27, 1635, and 1647-50, curtailing supply and forcing up oil prices in New Castile. ⁽¹⁹⁾ Similarly, there was an increase in oil consumption around 1640, even though the demand for wine fell rapidly. Apparently the olive groves planted when demand was strong began to produce just as war restricted exports. ⁽²⁰⁾ The decline of 1602-06, however, reflects the Court's absence from Madrid. ⁽²¹⁾ Underlying these variations is an upward trend in consumption from 1590 to 1620 that was reversed between 1625 and 1650, despite the brief upsurge in the early 1640's.

Eighteenth-century figures show a similar cycle of expansion and contraction, but much higher total consumption. By the 1750's, volume was double the seventeenth-century maximum and rose faster than population until 1792-98. With the subsistence crisis of 1799, oil consumption began to decline as population fell, and by 1807 had dropped to the level of the 1760's. It remained low until the 1830's despite population recovery, but after 1840 expansion was dramatic and in 1848 it was 50% above any previous maximum.
 
D. Meat
 
Meat includes several commodities, the most important being beef, mutton, and pork, but they cannot be sorted out before the eighteenth century. Pork is documented only after 1740, and beef and mutton were taxed together at different rates. Without knowing the ratio between the two, revenue figures cannot be converted to estimates of consumption. A few figures for beef and mutton consumption were developed from fragmentary seventeenth-century sources, and official totals exist for the eighteenth and nineteenth centuries (see Table 6.2 and Appendix D). In conjunction with revenue trends, they provide a profile of the volume and evolution of urban meat consumption. ⁽²²⁾
 
 

Pork consumption is documented for the later eighteenth century, and the expansion from 2.2 to 4.2 million pounds per year between 1740 and 1800 presents an exaggerated version of the concurrent cycle for beef and mutton. Annual fluctuations were sharper, and the trend clearly outpaced that of beef and mutton in the 1760's and '70's. The relative change is greater, as the quantity of pork consumed rose 75% between the 1760's and 1790's, and the subsequent decline is more precipitous. Apparently the rural economy could respond faster to the demand for pork, and for the urban consumer pork was cheaper if less desirable than mutton and beef.
 
E. Charcoal
 
The fifth mass-consumption commodity was charcoal, and the available annual figures are presented in Table 6.3. Consumption grew with population, but the timing suggests a supply problem and declining per-capita consumption during the 1790's. Annual consumption in the mid-1760's was 40% higher than the isolated estimate of 1695, and by 1789 it had doubled that of the 1760's--a rate of growth considerably greater than that of the population. Consumption then experienced an equally abrupt 30% decline in the 1790's.
 

[119] II. Per-Capita Consumption and Evolution of the Market for Agricultural Commodities
 
The preceding supply history allows a profile of urban consumption over time that is unique in urban history and permits a rough picture of overall market development in the course of 250 years. Using population estimates from Chapter 2, Table 6.4 presents the evolution of the city's per-capita consumption of agricultural staples. Where possible, consumption estimates are based on averages of adjacent years to minimize the effects of short-term shortage. The results are not directly comparable to the diets presented in Chapter 4, but they are remarkably close to those diet figures.
 

The evolution of wine consumption shows surprising elasticity of demand, given the conventional image of wine as a basic beverage, and also indicates the inability of wine-producing areas to supply large quantities at stable prices. The figure of 207 liters annually per inhabitant for the late 1590's is extremely high, but we must remember that most of the population was adult. The same is true for 1630, although by then average consumption is much closer to the 109 liters Bennassar found for Valladolid. These figures are easier to accept when compared to the 250 liters per-capita of Calvinist Geneva in the eighteenth century. ⁽²⁶⁾ Prior to 1610, wages in Madrid fell behind the price of wheat, and wine consumption grew more slowly than population or wheat consumption, as declining real income made it difficult for the poor to buy wine even at lower prices. After 1610 the situation changed as real wages rose--and even though the price of wine rose, that of wheat drifted downward and wine consumption soared. This coincides with the rise of the market-tax revenues shown in Chapter 2, a further indication of demand for less essential goods.

Thus wheat and wine consumption and population growth demonstrate the growing wealth of the early seventeenth-century capital. They also imply changes in the relationship of the Castilian hinterland to urban demand. [121] Wheat clearly became more available to Madrid after 1610, allowing the increase in real wages and in demand for other commodities. This is a significant rural response to urban demand, and as we will see, it reflects a restructuring of the regional economy. By 1630 the trends had reversed: inflation returned, real wages were declining, taxes were being increased, and bread prices were climbing. Wine prices, which declined in 1588-1611, soared to a high plateau that ended in a new upsurge in the 1640's, suggesting permanently higher costs in the industry. Individual wine consumption reacted to wine prices and began to decline as soon as the cost of a basic diet without wine began to rise faster than wages, after the prosperity of 1609-22. ⁽²⁷⁾ In the eighteenth century, wine prices began to rise sooner than those of wheat; but in the early decades, wine consumption may have benefited from the cheapness of wheat. As wheat became more expensive relative to wages, however, wine became less of a bargain. Around 1735 a worker could get seven pounds of bread by giving up an azumbre of wine, but by the 1790's he could get only four. Consequently, wine consumption rose slowly and erratically compared with population. Relatively high in the 1720's and early 30's, it fell as wheat prices jumped during the erratic harvests of the early 1750's. Wine consumption then oscillated around 500,000 arrobas per year to the end oí the century, despite the addition of 40-50,000 people to the city. Along with relatively rapid increases in the price of wine, this explains the decline in per-capita consumption after 1757. If we accept Madoz' estimate of consumption in 1850, ⁽²⁸⁾ the highest comparable figure for the 1840's is 65 liters annually per person, close to eighteenth-century averages.

Thus per-capita consumption of wine dropped sharply in the seventeenth century, hovered at 60 liters in the eighteenth century, with a decline at the end, and barely reached eighteenth-century levels by 1850. Compared with wheat, wine consumption was very sensitive to economic conditions. Whenever bread became expensive relative to wages, per-capita consumption of wine dropped. Urban population and aggregate wealth were thus less relevant to the wine market than were real wages and consumer preferences, and there was little in the structure of Madrid's market that allowed urban growth to encourage Castilian viticulture. It is a cliché that rigid and uneconomic land-owning arrangements stifled Castilian agriculture, but in this instance the nature of the market is an equally good explanation.

Consumption of olive oil did not react to the conditions that influenced wheat and wine, suggesting that its urban market was different. In the early seventeenth century, consumption grew regardless of changes in wheat and wine prices, and even despite increases in the price of oil, and demand remained [122] strong in the crisis-ridden 1640's and 1790's. This identifies olive oil as a commodity used primarily by the affluent, who were not forced by high price to surrender amenities from their diet. At the same time, olive oil shows a long-term trend the reverse for that of wine. From 5.3 liters per capita in 1630, consumption had tripled by 1769 and remained at 16 liters annually per inhabitant for the rest of the century. After a decline in the 1820's, it reached 25 liters per inhabitant in the 1840's, five times the seventeenth-century average. This does not mean that oil was widely used by the poor, since much of what entered the city was destined for street lighting and industrial use. The immunity of oil consumption to changes in the real wages of the poor reflects the fact that the poor could hardly stop using something that they made little use of anyway. Consequently, Madrid provided a sustained and growing demand for olive oil in contrast to viticulture. This represents an inducement to agricultural specialization derived from urban demand, but one that affected distant Andalucía far more than the Castilian interior.

Our calculations indicate that per-capita annual consumption of beef and mutton was relatively low at the end of the sixteenth century and had doubled by 1787. It declined subsequently, and as of 1850 had not exceeded the level of the eighteenth century. Use of pork followed the pattern of beef and mutton in the eighteenth century, but with higher rates of increase and decrease. At its peak, consumption of beef, mutton, and pork was about 4.2 ounces per day per person, a third of which was probably fat. This is a high figure for a pre-industrial city, ⁽²⁹⁾ but is not far from that Bennassar gives for Valladolid in the 1590's. ⁽³⁰⁾

If we accept that Spain was inherently better suited for livestock than other European countries, ⁽³¹⁾ the relative importance of meat in the diet is not an issue, and the trend of consumption is more significant. Since we cannot separate beef and mutton consumption, analysis of meat consumption is restricted. Its increasing availability in the urban diet is a clear reflection of the growing importance of stock-raising in seventeenth-century Castile. By the same token, the economic crisis of the late eighteenth century signaled a shift away from livestock, and per-capita meat consumption between 1787 and 1842 fell significantly before recovering its eighteenth-century level around 1850. The decline coincides with a steady expansion of rural population and agriculture. ⁽³²⁾ It is also interesting that the urban crisis in 1793-1807 [123] marked the beginning of a massive shift from mutton to beef in the diet. In 1796-99, beef represented 30% of the total of beef and mutton. By 1825, beef accounted for 60% of the total, and by 1847 it accounted for 77%. ⁽³³⁾ The recovery of per-capita meat consumption in nineteenth-century Madrid was clearly associated with a major shift in the nature of Spanish stock-raising.

The series on consumption of charcoal is consonant with other developments in the eighteenth century, although it is too short to provide much perspective. ⁽³⁴⁾ The high per-capita figure for 1789 is from a good source, and the decline in the 1790's is not an illusion caused by the documents. Charcoal was the household fuel, but it was also the source of industrial energy in the city. Bakeries, breweries, distilleries, chocolate makers, and metal foundries all used it extensively. In Chapter 5 we showed that Madrid saw a degree of elite-oriented industrial expansion in the 1770's and 1780's, and this accounts for the apparent increase in per-capita use of fuel. The economic difficulties that began in the 1790's curtailed development, as trade was disrupted and inflation caused the more affluent to cut back on marginal luxuries. At the same time, the drop in wages was bound to curtail household use of fuel.

Total and per-capita consumption of wheat, wine, and olive oil thus provide insights into three important elements of the Madrid market and show us how they changed with urban prosperity and growth. Wheat was essential to the poor at any price; hence urban growth brought increased demand, supply difficulties, rising prices in the city, and pressure on the real incomes of the poorer three-fourths of the population. Nevertheless, wheat consumption grew with the city, and sometimes faster, and every effort was made to mobilize supplies for the urban market. Wine was also a commodity of mass consumption, but was not essential in a supply crisis. As the real income of the poor declined, per-capita use declined, even though wine prices fell behind those of wheat. Thus Madrid's growth offered few market opportunities to viticulture and similar industries. Wine was a major agricultural commodity in the interior, but imperial and urban prosperity could not stimulate this important aspect of the rural economy. The urban elite used olive oil in increasing amounts, since they could afford any reasonable price without much concern for the cost of other staples. Since oil came from Andalucía, the stimulus of increased urban demand was transferred to suppliers outside of Old and New Castile.

Meat is usually presented as a commodity with an elastic demand curve similar to that of wine, and in some cases the two were affected by the same urban conditions. But in the Spanish context, the structure of the rural economy [124] was probably more important in determining levels of meat consumption. Despite widespread poverty, economic conditions allowed a level of consumption that was high for a pre-industrial city. Thus we see in urban meat consumption a reflection of the shift in Castilian agriculture toward higher-value commodities and reduced labor requirements that marked seventeenth-century Spain, as well as a reverse trend in the early nineteenth century.
 
III. Structure of the Urban Market
 
A. Agricultural Commodities, 1601-1800
 
The preceding discussion traced the evolution of effective demand in Madrid for basic commodities and showed that both urban income and rural conditions affected price and consumption patterns. The next step is to estimate the changing economic significance of the pattern of consumption for the rural economy and its elites. Our estimates of consumption and Earl Hamilton's prices provide an approximation of the annual value of each commodity, and thus a rough index of the gross return to suppliers. We can follow the relative significance of each commodity by comparing benchmark cross-sections. Figure 6.4 illustrates the relative importance of wheat, wine, olive oil, and meat in each of five periods, while Table 6.5 details the changes in volume and price in the early seventeenth century.

Expressed as five-year averages, the prices of these four commodities rose 30% to 19% between 1600 and 1631. After adjustment for the 31.2% increase in the price index, however, there was no increase in the real prices of wine and olive oil. Consequently, the gross return from those commodities could increase only as fast as consumption itself. To the extent that prices hid new taxes, the actual return to the producer rose more slowly than total consumption. Consequently, real increases in revenue in those trades could only come from greatly expanded volume or lower production costs. Given the decline of rural demand after 1610 and evidence of rising real wages in 1610-25, neither alternative is very plausible. Thus the fact that the consumption of wine and olive oil did increase in Madrid implies the decline of alternative markets and reflects the stagnation of trade at Sevilla and the collapse of Toledo and other interior cities. ⁽³⁵⁾

The contrast between the wine and olive oil markets and those for meat and wheat echoes the urban shift to more basic foodstuffs. It is worth noting that while meat consumption rose with population, mutton consumption apparently [125] rose faster than beef, reflecting a slower price rise. A declining Castilian labor force encouraged grazing, even as the difficulties of the European wool industry in the early seventeenth century sped more sheep to the slaughter-yards. Thus, even though its real price was static, selling mutton to the Madrid market remained a logical alternative for some land-owners, and the "decline" of Castile made another brief contribution to the general prosperity of early seventeenth-century Madrid.

The significance of this pattern is better appreciated if we look at the aggregate value of these four commodities together and examine their changing relative importance (see Figure 6.4). The decline in the relative importance of the wine industry is dramatic, the growing importance of meat is notable, and wheat came to represent nearly half of the value of the four basic staples. The real value of wine and olive oil sales increased by only two-thirds, while that of wheat and meat nearly trebled. Landed elements that could channel meat or wheat to Madrid apparently fared better than those providing wine or oil.

We cannot construct a similar cross-section of the commodities market for the later seventeenth century, but there is evidence for the direction of change. Wine continued to lose its share of what had become a generally contracting market as consumption continued to decline after the population stabilized. Wheat remained the most important commodity, and shortages were chronic despite the reduced urban market. Mutton had increased its share of the market in 1601-32, but as the wool market improved it became [127] more expensive relative to bread. ⁽³⁷⁾ Thus the poverty of most of Madrid's population continued to encourage the drift of Castilian agriculture toward wheat production and sheep, despite the decline of the overall market. The recurrent wheat shortages suggest that as urban demand fell, rural society moved toward local self-sufficiency and withdrew from distant markets. ⁽³⁸⁾ At the same time, the Madrid market supported grazing by providing an alternate source of income for sheep-raisers. By the late seventeenth century, the price of mutton in Madrid reflected conditions in international trade. When war disrupted the wool trade, the price of mutton fell in Madrid, suggesting a partial remedy for loss of access to wool markets. To the extent that Madrid thus encouraged export of raw materials, it reinforced the drift of the national economy toward dependence on European markets. ⁽³⁹⁾

The structure of the basic commodities market in the eighteenth century verifies and extends the declining relative importance of wine. From 30% of the basic commodities by value in 1632, wine fell to 25% in 1766 and less than 15% by 1800. Meanwhile, the volume of olive oil sales rose and its value increased from 2.8% of the commodities market to 10.5% by the end of the eighteenth century. Thanks to reduced urban demand, better crops, and lower prices, the relative importance of wheat declined in the early eighteenth century. By 1750, however, wheat consumption and prices were rising, and by 1800 wheat alone accounted for half of the market value of basic staples. Despite this growing emphasis on bread, the relative importance of meat remained much greater than in the early seventeenth century, reflecting the place of grazing in the structure of the regional economy.

The evolution of Madrid's market for agricultural staples reveals part of the urban-rural interaction centered on the city. As the city grew and acquired its distinctive economic structure, it reinforced seventeenth-century trends toward cereal monoculture, grazing, and local self-sufficiency in the interior, encouraging the separation of interior and peripheral economic life. Urban demand for wheat encouraged the search for new señoríos despite the rural crisis, while mutton consumption gave sheep-raisers an alternative source of income when the wool trade was disrupted. Viticulture became unprofitable as weak demand and scarce labor reversed the favorable conditions [128] of the sixteenth century, resulting in a long-term decline. As Madrid contracted after 1640, these processes did not simply go into reverse. Without the developed urban network of the sixteenth century, the simplification that marked the Castilian economy continued, and the emphasis on self-sufficiency, grazing, and grain was strengthened.

When Madrid began to grow again in the eighteenth century, it exerted the same pressures as in the seventeenth century, but was acting upon an economy already shaped by the first episode. Thus viticulture remained a stagnant industry, expanding slightly in the 1770's but faced with a static or declining market. Because of different market and production conditions, olive oil became substantially more important. Wheat consumption rose faster than population, as bread supplanted other foods in the popular diet because of declining real wages. By the last decade of the century, even meat, a basic staple in Madrid, was losing ground as a source of income for market-oriented agriculture. Thus as the city grew, the structure of income and demand in Madrid created a market for the hinterland that encouraged a limited range of market crops, especially wheat; while the growing demand for oil exemplifies the tendency of Madrid to transfer urban demand out of the city's Castilian hinterland altogether.
 
B. The Late Eighteenth Century: Staples and Other Commodities
 
If urban demand shifted toward one or two Castilian monocultures, it also gravitated toward more expensive commodities of distant origin, further contributing to Spain's economic dualism. This polarization of demand is apparent at the end of the eighteenth century, when we can document consumption, market shares, and geographic origin for a larger group of commodities. A detailed breakdown is presented in Appendix D, but the pertinent trends are summarized in Tables 6.6 and 6.7.
 

This information also allows us to refine the established simplifications about real wages in Madrid in the late eighteenth century by clarifying the sources of inflationary pressure. In the twenty years before 1789, the price increases were strongest among staples from the interior, but thereafter imports registered the fastest increases. Comparing money wages with general prices, Earl Hamilton suggests that real wages fell 13% from 1766 to 1789 and another 15% between 1786-90 and 1796-1800. In fact, during the first period the price of wheat rose 13% faster than the general index, that of charcoal 45%, and beef 20%. Pork, however, did not keep up with the general index, hence the relatively rapid increase in its consumption. Since these basic staples accounted for 70% of the budget of the poor, the deterioration of real wages prior to 1789 was far more serious than Hamilton implies. This makes the rapid deterioration of urban demographic conditions shown in Chapter 3 much more logical.

For the last decade of the century, Hamilton's figures show a 28% rise in the general price index and a 15% fall in real wages, but the impact of prices on wages is overstated. Semi-luxuries like soap, cacao, and sugar, which had barely kept pace with the general index before 1789, rapidly became more [130] expensive. The most notable increases were logged by overseas products and by olive oil, which outpaced the general index all through the last third of" the century. By contrast, wheat prices rose 20% less than the general index, and the same was true of charcoal. Beef also fell behind the general index, while mutton followed it closely, and wine continued its long-term decline relative to general price levels.

This confirms that even before 1789 the economic position of the poor had reached a point at which whatever happened to general prices, money wages followed bread prices because people were at the edge of subsistence. The misery this entailed is spelled out in the vital statistics of the 1790's, the subsequent decline of total population, and an urban market in which consumption of the cheapest staples (wheat and pork) rose faster than the population, while demand for other commodities used by the poor stagnated or declined.

In the third quarter of the eighteenth century, therefore, urban growth brought greater demand for staples from the Spanish interior, even though it did not create a diversified market for that hinterland. Charcoal, pork, beef, wheat, and wax registered 25% to 90% increases in volume and 45% to 175% increases in market value when adjusted for inflation. Around 1789, however, urban demand shifted to the cheapest staples on one side and semi-luxuries on the other. In the last years of the eighteenth century, therefore, Madrid bought a static or declining volume of hinterland products, except for wheat and pork. Adjusted for inflation, the market value of hinterland commodities consumed in Madrid fell anywhere from 40% in the case of charcoal to 5.5% in the case of wheat. By contrast, consumption of sugar, olive oil, fish, and cacao all rose noticeably, and their adjusted retail values climbed even more. Despite higher international prices, the city increased its consumption of these semi-luxuries, a sure indication of an increased polarization of urban incomes and of the dualized structure of the urban economy. The implications of this for a landed elite with urban habits are obvious. ⁽⁴⁰⁾
 
C. Structural Trends from Eighteenth to Nineteenth Centuries
 
Fragmentary sources make it difficult to follow these trends into the nineteenth century, but some changes are apparent (see Table 6.8). The first section of this chapter and Table 6.4 on per-capita annual consumption offer figures that provide a starting point.
 

It would be surprising if the trade structure of any major pre-industrial city did not include this pattern, but Madrid's failure to draw manufactures from its hinterland is striking. Moreover, there are signs of a progressive sharpening of this commercial duality. This is the opposite of what happened around cities like Barcelona, London, or Amsterdam. From its emergence as capital, Madrid drew to itself the skilled crafts of the interior and developed elite tastes and an income structure that restricted urban-rural economic interaction of the sort that stimulated complex hinterland economies elsewhere.

Thus far, most of the evidence for this trend has involved the impact of urban growth on demand for foreign or domestic agricultural commodities. Comparison of urban imports in 1780 and 1847 illustrates the same shift in the origin of industrial products and raw materials. Demand for nonessential goods from distant suppliers substantially surpassed eighteenth-century levels and rose faster than at any previous time (see Table 6.12).