[338] I. Occupational distribution of the economically active population
 
The principal sources for the occupational structure are two versions of the Catastro of 1757. One was published by Antonio Malilla Tascón and compiled in 1757. The other, compiled in 1770, was found in the municipal archives. The two are structured differently and were drawn from the same missing data bank for different purposes. The earlier version was drawn up as a report on the aggregate income of the city for use in general calculations of the ratio of national income to royal revenues. This was the first step in developing quotas for a new single tax. The version of 1770 was developed in connection with the effort to actually implement the tax and assign obligations in conformity with the rates designated for different types of income. Neither source was constructed with the objectives of this analysis in mind, and both reflect the fiscal concerns of the project for which they were compiled. As a result, aggregate income figures are attributed to many corporate groups, royal offices, and guilds without showing internal distribution. Much of this more detailed data was presumably available in the original supporting data, and a good deal of it came out in the comparison of the two versions. Nevertheless, it was sometimes necessary to make tentative estimates of the numbers of individuals and of individual incomes. Where this was done, it was by analogy with comparable cases where the data was less ambiguous.

The two most difficult occupational groups to estimate were the bureaucracy and the Church. For the bureaucracy, incomes were listed by bureau or administrative unit. Here it was necessary to go outside the two versions of the Catastro. The census of 1787 gives total royal employees at that time, while the Guia de Forasteros allows a count of the relatively important officials in the city. By counting the latter, and comparing the number with royal [339] employees recorded in 1787, a ratio was established which allowed an extrapolation from the number of officials in the Guía at the time of the Catastro to an estimate of government employees at that time. This is the source of the estimate of 3,000 inserted in Table 4.1. While the procedure is less than ideal, the margin of possible error is not large enough to affect the overall structure that emerged. The one part of the government sector that could not be checked was the membership of the royal family and its dependents.

The religious personnel of the city presented a different and partly semantic problem. Contrary to Old Regime tradition, adult religious personnel have been classified as part of the active population. This is done on the grounds that they were performing spiritual functions (and often economic and social ones) valued by the society. Indeed, monasteries and convents ran bakeries, wine shops, schools, orphanages, and hospitals. Analysis of the city census that accompanies the 1757 version of the Catastro, combined with lists of individual incomes received by segments of the religious personnel of the 1770 version, allowed a fairly complete count.

Lesser problems emerged in the enumeration of staff members of religious bureaucracies and of the nonprofessional employees of the medical and legal professions and the Five Greater Guilds. The 1770 version allowed separation of the aggregate incomes of ecclesiastical bureaucrats, doctors, lawyers, and merchants from the aggregate income of their employees, and allowed some occupational classification as well.

The sector for which numbers of economically active persons could not be estimated separately included educational institutions, hospitals, and hospicios. The institutional occupants were not economically active, but teaching and administrative personnel obviously were. Since these institutions were tied to Church and state by both personnel and endowment, it was assumed that their active personnel was accounted for in one or the other. In cases such as company and agency chaplains, or teaching and nursing nuns and monks, they were included with other members of corporate entities.

Classification of the food, construction, and manufacturing sectors was generally straightforward. The 1757 version of the Catastro gave considerable detail on active individuals and individual wages. The servant class posed a more difficult problem. In one place there is a single entry for 8,000 sirvientes of all types, with an aggregate income figure. This appears to include servants with long-term contractual employment. The tabulations also show a group of some 3,000 "gente de librea," who are distinct from the day-laborer category. This phrase is associated in other sources with lacayos -- lackeys, grooms, footmen, and other manservants. Listed with a variety of daily rates of pay, these "men for hire" are classified separately from construction labor, commercial porters, and servants with long-term contractual arrangements. By a process of elimination, these men-for-hire have been [340] defined as casual servants and grouped with the criados and sirvientes as part of the service sector.

The concrete numbers in the sources and the guesses, estimates, and extrapolations mentioned above were used to construct the figures summarized in Table 4.1. The estimated total active population thus developed could be used to establish the percentages given in Table 4.2.
 
II. Distribution of income among the economically active population
 
The income attributions in Table 4.5 reflect the same estimates and approximations used in establishing occupational distribution. The main part of Table 4.5 shows those income attributions which are given in one way or another in the two versions of the Catastro. In some cases, as suggested in the preceding section, this involves use of income figures in analogous cases to estimate the numbers of persons and average incomes in groups such as the dependents of professional and business enterprises. The largest aggregate which could not be broken down was the 8,000 "sirvientes de todas clases," with the result that the mean income of that large group is shown. The effect is no doubt to reduce the relative size of the 0-1000-reales income group, since it is hard to believe that most of the women servants were paid anything like the mean of the group. The relatively high average income of the servant group reflects the way in which annual incomes were calculated. Artisans, jornaleros, and gente de librea were attributed with 180 days of work per year. Criados were assumed to work 250 days, and meals were included as part of salary. As a result, even a relatively low daily wage produced the impression of relatively high annual incomes. To the degree that these conventional assumptions of the time were accurate, it appears that household service was far from the least attractive economic niche in the city.

In the case of propertied and titled persons, the listing in the main body of the table is based on the city-derived incomes attributed to them by the compilers of the Catastro. That same group is repeated at the bottom of the table, but redistributed according to the adjustments mentioned in the text for income from outside the city.

Also at the bottom of Table 4.5 is a tentative income distribution for government personnel. It is based on a survey of the Guía de Forasteros, available detail on agencies given in the version of 1770, and indications of pay and numbers of subalterns in various sources. It may reflect overestimates in the middle ranges, but salary figures for public and large private organizations from later periods suggest that even relatively minor employees got a "living wage" from the Crown -- a fact which is confirmed by the demand for such positions.

[341] Many professional groups appear with corporate income attributions that could not be disaggregated, and it was necessary to settle for the mean incomes in each group. Where possible, distribution of incomes among their employees was estimated by analogy with similar personnel shown individually in the records. Similarly, distribution of endowment income within many institutions and offices could not be estimated, and it was necessary to resort to averages.

A preliminary comparison of the income figures in the two available versions of the Catastro showed a number of discrepancies which suggested that the 1770 figure had been updated. A careful analysis of the figures for a number of guilds revealed that in fact the data base was identical, but that the calculations for 1770 were somewhat less precise. The early version, for example, listed widows of masters with shops separate from masters with shops, so that the widows could be shown with a commercial income but without a daily wage from industrial activity. The later version simply assumed that all masterships received daily wages as well as commercial profits. This type of discrepancy accounts for virtually all conflicts between the two sources on this point; and where possible, calculations were based on the more detailed data of the earlier version published by Malilla Tascón.

In general, most income attributions that had to be estimated or extrapolated are biased upward when any uncertainty is involved. One of the points of the analysis is to show the extreme inequality of income distribution in Madrid, and any upward bias works against that hypothesis. Thus such inequality is almost certainly somewhat less extreme in the findings than in reality. Table C.1 presents a more detailed version of the distribution of income among occupational groups than could be fitted comfortably into the text.
 

III. Average annual income in occupational groups
 

The following tabulation shows the composite annual income, including wages and commercial and property income, for most professional and skilled groups in the city -- officials, doctors, guild-masters, etc. It is not a complete listing of economically active persons and does not include the incomes of subalterns or wage-earning employees in these groups. Government officials excepted, it illustrates the relative income levels of various occupations in more detail than was possible in the text. It also illustrates the nature of the urban market by documenting the wide range of special occupations and the small number of practitioners in each specialty. This speaks to the narrowness of elite demand in the city and the lack of industry oriented to export from the urban economy.
 

To supplement Table C.2 and provide a fuller sense of urban income distribution, Table C.3 indicates the annual incomes attributed to a number of unskilled types of employment in the city.

Table C.4 elaborates further on income levels, showing the range of wages earned by employees of guild masters in the city. The extreme range of the oficiales or journeymen reflects (a) the inclusion of a number of highly skilled journeymen in precious-metal-working and (b) individuals in some royal establishments who obviously were functioning as foremen or overseers.
 
 

Without a measure of the value of the goods involved in Madrid's trade, the elaborate contemporary list of urban imports and consumption compiled in 1789 is of little analytic value to us. Recognizing that at best any such calculation can yield but a rough approximation distorted by unknown product variety, variations in measurement, and outright fraud, an attempt was made to attach prices to as many imports as possible. Our preferred source was Earl Hamilton for the late 1780's, since the commodity prices he gives are based on the market in Madrid. Where possible, these were compared with other sources, including some of the commodity prices given in the work of Gonzalo Anes, which come from different documents and have proved generally reliable. Prices for foodstuffs of all sorts were found in the lists of price ceilings published in the Correo Mercantil during 1793. Hamilton's figures suggest that prices did not differ greatly between 1789 and 1793, so that the ceilings given provided a rough approximation. In addition, the Correo Mercantil published the current prices of several commodities in various ports. The prices of a number of urban imports were established by taking quotations from the most obvious port of entry or point of origin for goods en route to Madrid. Such figures are unavoidably low, since they do not include the cost of transport and handling en route to Madrid. In some cases, prices for manufactures could be estimated by reference to the Balanza de Comercio of 1795, although these suffer even more from various deficiencies. Many commodities that could not be priced directly were close substitutes for those that could be priced, and in those cases the prices of the substitutes were used.

Throughout these calculations, three biases were unavoidable, and therefore they were handled in a consistent way. Food prices from 1793 are assumed to have a slight upward bias relative to 1789, although they tended to include processed goods and luxuries which were relatively expensive anyway. The goods priced through port-city quotations were colonial products or [348] metals, and the bias is clearly downward in those cases. In the case of goods which appear to have been near substitutes, the prices which were assigned are probably lower than the missing prices they replaced. The effect of these distortions is a slight overstatement of the value of subsistence commodities, most of which came from the Castilian interior, and a definite understatement of the value of colonial and other imports which were seldom part of the subsistence commodities that everyone needed. As a result, the proportionate value of the city's imports that were luxury-oriented or nonessential manufactures and raw materials is understated.

About 75% of the goods recorded have been priced, and the estimate that this represents 75% of their value is very rough and tentative. More than 25% of the named items could not be meaningfully priced, but a great many of them were recorded in small quantities. The most serious omission, for which no correction could be made, was the price of building materials. Bricks, plaster, stone, and lumber simply did not show up in any sources. Wood of some types did appear, but the tremendous variety in wood types, qualities, and units of measure made it necessary to abandon the few available quotations. Thus the totals cannot be used for precise calculations, but they do establish the magnitude of values. This magnitude fits well with other figures on the urban economy, and the internal proportions of the total are likely to be fairly accurate.