The 1770s, a lively decade for quality control: the case of the Royal Tobacco Factory of Seville

1. Introduction

As Miller (1994) points out, management and accounting can only be understood in an organizational and social context. Moreover, following Carmona et al. (2004) this understanding is improved if, from a historical perspective, we analyse the accounting change episodes over the time. Following these premises many academics have been pushing the boundary of management and accounting to focus on its relationship with the organizations and society at large, in the last couple of decades (Macintosh & Hopper, 2005, p.1).

Since the 1990s, management accounting historiography maintains the relationship between accounting and effective management (Fleischman & Tyson, 1993). This relationship has been confirmed by recent studies partly because of the explosion of accounting history research in the last few years (Carmona et al., 1999; Carmona, 2004). Therefore, several studies have demonstrated that this relationship existed even before the Industrial Revolution (see Hoskin & Macve, 1988; Fleischman & Parker, 1990,1991; Edwards & Newell, 1991; Carmona et al., 1997, 1998,2002; Prieto & Larrinaga, 2001; Núñez, 2002a,b; Carmona & Donoso, 2004; Carmona & Gómez, 2002). On the other hand, the literature has shown that the development of management techniques in eighteenth-century Spain was similar to that in other advanced countries like the UK (Gutiérrez et al., 2005).

More concretely, several case studies have addressed management techniques in specific eighteenth-century organizations, for example, the management of operational time (Gutiérrez & Romero, 2001) or quality management (Cox, 1990; Álvarez et al, 2002). Cox (1990) studies quality in Josiah Wedgwood's pottery firm between 1730 and 1795. Álvarez et al. (2002) analyse the use made of accounting calculation techniques as a contributory element to quality control in the case of the Royal Tobacco Factory of Seville (RTFS) during the second half of the eighteenth century using information obtained from the instructions given for the management of the Factory. In this sense, the present work extend the one by Álvarez et al. (2002), which presents the role of the instructions of the Royal Tobacco Factory with regard to quality.

The objective of the present work is to show several examples of quality decisions taken ad hoc and the reasons why quality was taken into consideration for management purposes in the RTFS in the second half of the eighteenth century. We will see that Tobacco Factory managers considered quality when they took some ad hoc decisions in different phases of the production process. In these cases, the quality measures were aimed to improve customer satisfaction and indirectly to improve the royal income from tobacco taxation. Besides, we will analyse how attention to quality was driven by two forces: on one side the Crown's desire to raise the royal income, and on the other, a will on the part of the managers of the RTFS to rationalize the production process according to the ideology of the Spanish Enlightenment. That is, the public happiness of the population engendered through the rise of the royal incomes.

The data for this research has been gathered from primary sources available in the Altadis Historic Archive (Archivo de la Fábrica de Tabacos de Sevilla, AFTS), National Historic Archive (Archivo Histórico Nacional, AHN), and General Archive of the Indies (Archivo General de Indias, AGI).The next section of this article introduces the context of the second half of the eighteenth century, and section three describes the RTFS and its production process. Section four presents seven decisions aimed to improve quality, or to improve the efficiency of the production process with implications for the quality of the final product. We then discuss the quality measures in the light of the eighteenth-century context and present our conclusions.

2. The second half of the eighteenth century in Spain

At the beginning of the eighteenth century, when Carlos II, the last Hapsburg monarch, died, a new dynasty came to Spain, the French Bourbons. The new dynasty meant to introduce governmental administrative reforms in response to the bureaucratization and bad administration of the Hapsburgs over the previous two centuries (Vicens Vives, 1987). The dynastic change is associated with innovations in social and economic politics inspired by the French Enlightenment model, which generated reforms that spread to Hispanic dominions overseas. The first half of the eighteenth century established the basis from which the most ambitious reforms were to be undertaken from 1760 onwards. The dynastic change diffused Enlightenment ideas throughout Spain and gave rise to a new regulatory framework dominated by the principles of rationality as a means for seeking the general objective of the public happiness (Sarrailh, 1992).

To reinforce and modernize the state, it was necessary to solve the financial problem of the Royal Treasury, badly managed and spent in fruitless religious wars by the Hapsburgs.1 One solution was to increase the public revenue (Artola, 1982, p.254) through the improvement of collection while trying not to create new taxes (General Archive of the Indies, AGI, Mexico 1249). The new administrative system was created and guided by a set of well-trained and competent public officials in the king's direct service called Enlighteners (Cespedes del Castillo, 1983).

Throughout Spanish politics, reforms were based on the wish to develop production and commerce, to shape a solid national economy, and to protect it against foreign competition (Anes, 1994). For these purposes the reformers used, among others, two apparently disconnected instruments: education and royal factories.

Education

The reign of Carlos III (1759-88) coincided with the height of the Spanish Enlightenment, when the intellectual figures and politicians joined forces. One of its main objectives was the education of the population (Sarrailh, 1992). To this end the Enlightenment used three instruments: the Economic Societies of the Country's Friends (Sociedades Económicas de Amigos del Pais), University reform, and the promotion of non-university technical centres.

The Societies of the Country's Friends were one of the most representative instruments of the official Enlightenment in Spain. The first one was founded in the Basque Provinces. Its success in fostering economic development pushed Campomanes to promote them from the Council of Castile, so that from 1774 to 1804 more than 50 societies were founded throughout Spain. The Seville one was founded in 1775 (Alvarez-Dardet et al., 2003) - as we explain below, just before a peak in reform at the RTFS. These societies tried to overcome educational deficiencies. Concretely, their mission was to keep their members knowledgeable about the economic and social situation of their territory, to give professional training to peasants and artisans, and to project and sponsor regional or national economic reforms. They were joint enterprises between localities and the state in which a part of the educated and leading class put itself at the service of the official reform programme (Anes, 1994; Fernández, 1994).

These societies joined Enlightenment adherents with the objective of discussing and diffusing technical improvements and scientific advances. Usually each society had three commissions whose members were recruited according to their specialization and knowledge. The first studied agriculture; the second mechanical tools and craftworks; and the third, industry, commerce, and navigation. The societies gave annual awards as incentives for technological and scientific improvements (Álvarez-Dardet et al., 2003).

University reform began after the expulsion of the Jesuits (by a decree signed by King Carlos III, 27 February 1767), which pushed the government into intervening in educational matters, as the order had had a preponderant role in advanced education (Domínguez Ortiz, 1988). The king commanded the Valencian intellectual Gregorio Mayans (Domínguez Ortiz, 1988) to prepare a suitable new syllabus for all the Spanish universities, which was presented the same year.

At the same time, other personalities of the Enlightenment also received similar commissions for other universities. One of them was Pablo de Olavide for the University of Seville. Olavide drew up a plan for the reform of this University in which he demanded the secularization of teachers, the bringing up to date of the subjects, and a pedagogical renewal with the establishment of textbooks under state control.This plan inspired others, such as the one at Osuna University (López, 2004).

Moreover, the administration also promoted non-university centres for higher education, such as the Royal Nobility's Seminar of Madrid, the Royal Study of San Isidro, the School of Mineralogy of Madrid, the Asturian Institute of Mines, and the Royal Study of Machines. These centres, mainly oriented to specific issues, encouraged the creation of skilled labour (Fernandez, 1994) by spreading a training network throughout Spain. These practical studies made easy the general rationalization of the production processes and the improvement of the quality of the products in the royal factories.

The Royal Factories

Enlightenment economic reforms sped up after 1759, with the support of Carlos III. Some of these reforms were aimed at free trade, while others, such as the creation of royal factories, were aimed at the protection of national industries. This era was distinguished by some economic success, although it came to an end because of the French Revolution by 1790.

In the previous centuries, entrepreneurs had failed to manufacture enough goods for Spanish and colonial consumption, to the benefit of manufacturers of other European countries. In order to halt the industrial dependence of Spain, the Bourbons created royal factories following the lines of French Colbertism (Helguera, 1991; Fernández et al., 1992), allowing them to import foreign machinery (Anes, 1994).

Apart from the objective of producing competitive goods for both the trade with the colonies and the Iberian market, the royal factories accomplished at least three further purposes. They were conceived as technical institutions for the dissemination of industrial knowledge and skills (La Force, 1965). second, even though early in the eighteenth century royal factories were compatible with guilds, after the liberal reforms started under Carlos III they were used as instruments to destroy guild privileges. Finally, in the case of some monopolies, some royal factories were also devised to maximize public revenues. Among the factories of the third group was the RTFS.

3. The Royal Tobacco Factory of Seville

The RTFS during the second half of the eighteenth century, has been studied in detail for the from a historic perspective by Oliver (1987), Rodríguez (1990,1992,1993, 1994) and Morales (1991), and from a management and accounting perspective by Gutiérrez (1993,1999), Carmona et al. (1997,1998,2002), Romero (1997), Gutiérrez and Romero (2001,2005) and Gutiérrez et al. (2005). From a few years after the discovery of America, tobacco came to Europe through the Seville harbour. Therefore this city was the first European one where tobacco was processed. In 1636, a tobacco monopoly was established in Spain, comprising the import of raw material, the manufacture of tobacco products, and their distribution in Spain and eventually abroad (Goodman, 1994; Comín & Martín, 1999). Thus, almost from the beginning, the production and sale of tobacco were not free for private individuals. This was because the Finance Ministry soon recognized that it could get significant resources from taxing tobacco, even though this meant the appearance of a parallel black market (Rodríguez, 1994). In the seventeenth century, the tobacco monopoly was rented to private individuals, though the records show a failed attempt by the state to take over direct control of the monopoly in 1684-5 (Rodríguez, 1990).

The administrative reforms of the new Bourbon dynasty included the creation of a Tobacco Agency that, from the beginning of the 1730s, reported to the Finance Ministry. The Tobacco Agency was a public institution that dealt with tobacco production, distribution, and tax collection. The tobacco tax was one of the most important revenue sources for the Crown in the eighteenth century (Artola, 1982). Therefore, the management of tobacco manufacturing was closely followed by the Ministry of the Royal Treasury and even the king himself.

At first, the only factory authorized by the Tobacco Agency was located in a small building in downtown Seville named the Fábrica de San Pedro (Factory of Saint Peter). As demand skyrocketed, the factory quickly ran out of room. Gutiérrez (1993) documents that the demand increased from a little more than 1,000 pounds in 1700 to the century's maximum of nearly 4,000 in 1737 - a rise of 271 per cent. Accordingly, the factory spread out into other houses around the first building.

The enlargement gave the factory's internal layout an irregular outline, making the production process and internal control difficult. In these premises, the managers could barely carry out production programmes and usually launched products onto the market in a great hurry, so attention was not paid to quality, and the finished products were full of defects (Rodríguez, 1994).

To solve this problem, the Tobacco Agency built a new factory (La Nueva Fábrica) outside the city walls, and this was finished in 1770. The design of the New Factory was aimed at fulfilling the demand for tobacco at that time. The regular building, formed by a network of columns and vaults, made it possible to separate workers in different departments and prevent movement among them, facilitating control and avoiding smuggling and inefficiencies (Carmona et al., 2002).

The Tobacco Agency enacted regulations (Instrucciones) to standardize the production process, the internal organization of the RTFS, and the responsibilities of the different posts in the factory (Romero, 1997). These rules, periodically read aloud to the workers, defined the best know-how for tobacco production as well as the specifications of the finished products (Álvarez et al., 2002). In general, if there were significant changes in the organization or product process of the factory, there were new instructions (Romero, 1997). The production process in the Factory of San Pedro was regulated by an instruction given in 1744, which treated specific problems in the production process, such as the night shift or product quality, and was aimed to improve the efficiency of the installation (Romero, 1997; Álvarez et al., 2002). There were instructions for the New Factory in the years 1761, 1779 and 1790.

The 1761 Instruction, signed by the enlightened Treasury Minister Esquilache, specified the new production process, detailing responsibilities and introducing the best know-how for tobacco processing, as well as a new accounting process (Carmona et al., 2002; Álvarez et al., 2002). The new accounting process meant a significant rise in the number of accounting books and statements, from 33 to 60, and a sophisticated flow of information inside the factory and between the factory and the Tobacco Agency (Madrid). This sophisticated accounting system allowed control of products and materials throughout the production process (Carmona et al., 1998, 2002). With the New Factory it was possible to improve other aspects of the management, such as operational time (Gutiérrez & Romero, 2001) and the quality of the finished products (Álvarez et al., 2002).

Production capacity changed significantly from the middle of the eighteenth century. The demand for snuff was falling while that for cigars was rising: in 1740 the snuff share had been 68 per cent and the cigars share 32 per cent; in 1770 the figures were 52 per cent and 48 per cent respectively (Gutiérrez, 1993, p.116). Product mix became problematic because, traditionally, snuff had been the most important product and even the New Factory had been designed to produce mainly snuff.

As Table 1 shows, from the 1770s demand for snuff tended to be stagnant and slightly decreased (Rodríguez, 1992, p.165). This meant that once the productioncapacity problem of the Fábrica de San Pedro had been resolved and it was no longer necessary to use a night shift or skimp on product processing time, it was possible to confront the problem of quality.

Two outstanding figures of the Spanish Enlightenment during the 1770s, Carrasco and Jovellanos, were named to manage the New Factory. Both of them had longstanding communications with the Treasury Minister and besides were founders of the Economic Society of the Country's Friends of Seville.

4. The production process

The Havana Company, partly owned by the Spanish Crown, imported tobacco leaves into Spain from the Americas. Upon their arrival in Seville, the raw materials were transported to the premises of the RTFS. The tobacco leaves were checked and classified at their entry into the factory for points of origin, quality, colour and size (Rodríguez, 1993, p.184).

The production process was performed in the RTFS, although there was a small cigar factory in Cadiz, which was dependent on the one in Seville during the second half of the eighteenth century. Likewise there was a small factory in Havana, which was managed by the Tobacco Agency and made an intermediate product known as polvomonte de indias or snuff partially processed in Havana (Rodríguez, 1993, p.112).

When the Crown took direct control of the factory in the 1730s, the factory organizational structure was simple, comprising a superintendent, only one accountant, and only one trustee. In contrast, in the New Factory the organizational structure was sophisticated, with six accountants and five trustees coordinated by two production managers (Carmona et al., 1997).

In the snuff factory, operators were lured daily at a fixed wage according to then professional category. In general, workers in the snuff factory required physical strength and a certain skill for operating mills and tending to the horses used in the mills. The process of manufacturing the snuff powder was carried out in seven stages referred to by old Spanish names (see Figure 1): azotea (sun-drying), monte (first milling), moja (mixing), entresuelos (airing), repaso (second milling), fermentación (packaging), and distribución (distribution) (Morales, 1991).

The manufacturing process for cigars was based on handwork, as each operator carried out the entire job of rolling the tobacco leaves into cigars according to predetermined conversion rates. Operators were hired on a weekly basis. The salary of each operator was determined by a piece-rate system that requested at least a target number of cigars of a set quality. If the workers did not produce cigars with the quality required, they had to remake them (Carmona et al., 1997).

Once completed, snuff and cigars were sent to a distribution network coordinated by the Tobacco Agency and formed by regional branches located in the main Spanish cities (Carmona et al., 1998). Each of these had a warehouse and a responsible clerk. From there the tobacco was distributed to the selling points. The Tobacco Agency collected the income and transferred revenue to the Crown Treasury (Carmona et al., 1998).

5. Quality in the RTFS

Quality was generally understood as resulting from the best production methods, and the instructions gave specifications for each product. However, given that tobacco is a living plant, many circumstances have an influence on the finished products. Therefore the production manager, who was a specialist, had leeway to make quality decisions throughout the production process. The 1744 Instruction embraced quality as follows:

The practice to be adopted is for the leaf to be exposed on these sunning roofs not at times of greatest heat but when it is more moderate, taking the greatest possible care that the exposure should be precisely limited to achieve the required juice/liquor content, avoiding the well-known problem of excessive evaporation... By following this rule, a good quality of tobacco will be assured during the [production phase]. (AFTS, 1744, 2 Instruction)

The interest taken in quality in the 1744 Instruction was due basically to the lack of space and high demand, and the consequent defects in the finished products, as the following passage shows:

One of the main reasons why tobacco has up to the present time lost much of its quality and good taste has been the malpractice of putting water and earth in the troughs for three and four days, without concern for what reason and experience demonstrate; this is that water and earth left for such a length of time go rotten and cause a disagreeable smell, which definitely ruins the tobacco... (1744, 2)

The 1744 Instruction also emphasizes the inspection and classification to which the tobacco leaf should be submitted on arrival at the factory, describes in detail the activities that must be performed at each stage of the snuff production process, and stresses the need to maintain order and cleanliness in the workshops.

Although staff and workers at all organizational levels were involved in the achievement of quality, the stage of the repaso, or second milling of the snuff powder, the most important stage for the appearance of the final product, was the responsibility exclusively of the superintendent himself and the senior managers (1744, 6). A Royal Order dated December 20,1730 said expressly that "the quality of the tobacco of Seville must be the best produced in Europe','and the last words of the 1744 Instruction specifically give the superintendent responsibility for quality:

[The Instruction] safeguards the best quality of the tobacco that ought to be the main objective of the Superintendent. (Instruction 1744, page F)

Nevertheless, by the middle of the 1770s the quality of the products still was not at all acceptable. The quantity of raw material stored in the factory was excessive; but worse, most of the stocks were ruined and unusable. The rolled cigars were without bouquet or juices, were not tightly rolled, and were found to be full of defects.

Faults were also found in the processing methods (1779, "Instruction to be observed by the operations managers and their corresponding responsibilities','6). Therefore, a Commission was sent to the RTFS with the objective of improving the quality of the production process (Carmona et al., 1997). This commission attempted to implement measures to correct these defects and prevent their reappearance in the future.

As far as quality is concerned, it was ruled that the production manager, who had the ultimate responsibility for production activities, should make the inspection of the final product (previously, the superintendent had performed this function) (see Figure 2). With this measure, the Tobacco Agency managers linked greater specialization in the managing of the production process with the quality of the finished products. Also, new rules were issued for the establishment and functions of a new operations management post related mainly to quality inspection, that of a second production manager who was to assist the senior one (AFTS, 1779 Instruction: "Standard to be observed by the operations managers and their corresponding responsibilities,"8).

The quality of the finished products was a matter of concern for the superintendent, the Tobacco Agency Board, and the finance minister as well as for the king of Spain himself, as will be shown in the following seven cases. These cases show measures that were taken ad hoc with regard to quality and demonstrate that quality was taken into consideration from the beginning to the end of the production process.

Case 1: decisions about the polvomonte de Indias

The tobacco named polvomonte de Indias was a kind of snuff partially processed in the Havana factory (Cuba) and finished in the RTFS. The phases of azotea (sun drying of tobacco leaves) and monte (first milling) were performed in Havana, whereas moja (mixture), entresuelos (airing), and repaso (second milling) were performed in Seville (see Figure 1). Then the snuff was canned, distributed, and sold in Spain and, occasionally, abroad. Imports of polvomonte accounted for 30 per cent (measured in physical units) of total inputs of tobacco coming from America (Rodríguez, 1993, p.239).

In the middle of the 1770s there was a large quantity of tobacco leaves stocked in the factory stores. Therefore the superintendent asked the finance minister to decrease the imports of polvomonte by sending less tobacco from Havana (AFTS, leg. 194, letter dated 21 October 1777). At the same time, the superintendent also complained to the minister about the bad quality of the polvomonte, which was mixed with other materials like sand and small sticks, mainly owing to the negligence of the workers.

The problems with the bundling of polvomonte continued ... thus, five bundles coming from Havana have brought a great amount of sawdust, manure, sand, and small sticks. (AFTS, leg. 194, letter dated 21 October 1777)

As a consequence, the finance minister suggested:

... the necessity of sending to Havana an expert of the factory, who, knowing the tobacco that existed [and was needed] in Seville, would supervise the shipments made from that moment on. (AFTS, leg. 194, letter dated 21 October 1777)

Besides, and more importantly, this expert would examine the entire production process of polvomonte made in Cuba, from the agricultural conditions of tobacco farms through the selection of tobacco batches to be shipped to Seville (AFTS, leg. 194, letter dated 21 October 1777).

Antonio Aguilar de Zelas, an official expert in tobacco recognition, was appointed for this task and travelled to Havana. As the minister of finance had suggested, the visit was urgently needed to tackle the existing problems. The cost of sending this expert from Seville to Havana included the journey, the cost of a servant, and the annual salary of 30,000 reales (Spanish coin of the time).2 This was a well-paid job; by way of comparison, the superintendent, the best paid person in the factory, earned 33,000 reales per year, or the equivalent of the average salary of 18 workers in the snuff workshops of the RTFS (Gutiérrez, 1993).

Case 2: decisions about the almagra

The almagra was a type of red sand that was used to give the appropriate colour to the snuff in the third phase of the production process (moja, or mixture) (see Figure 1). In order to improve the quality of the snuff, the director of the Tobacco Agency ordered an experiment to test the influence over the colour of the snuff of five different kinds of almagra obtained from five different mines. As a result of the experiments, the agency made a quality ranking of the different types of almagra with the purpose of encouraging the use of the best one.

We have checked the five different samples of almagra with the assistance of the operations manager. After several experiments, we have found that, with the exception of sample number one, which gave too light a color to the tobacco, the other four could be used in the factory. Besides, we have noticed that the number three is the best one. (AFTS, leg. 194, letter dated 3 February 1777)

On the other hand, the experimenters also noticed that the rain impaired the quality of almagra during its transportation to Seville. The almagra was shipped from Mazarrón, a Mediterranean coast town about 500 km from Seville. It was decided that the transportation from Mazarrón to the RTFS should be accomplished in summer (the dry period) to improve quality,3 in spite of the larger stocks that this decision required (AFTS, leg. 194, letter dated 26 February 1777). The superintendent of the Mazarrén mines wrote:

I think that you should determine the needs for almagra for the factory by next May, so as to dig out in June and ship in July and August to avoid having the almagra ruined by the rains as in the last shipment. (AFTS, leg. 194, letter dated 26 February 1777)

Case 3: decisions about tobacco transportation

Since 1717 the tobacco leaves and polvomonte de indias had arrived in Spain through the harbor of Cadiz (Aguilar, 1989, p.390), which was granted the Monopoly of Trade between Spain and its overseas colonies until that trade was deregulated in 1778. The raw materials were then shipped from Cadiz to Seville up the Guadalquivir River in small boats, because navigation throughout this river required ships with less draught. In return, the finished products that were sold in the Cadiz area were also transported down this river from Seville (see Figure 1).

The superintendent of the RTFS complained to the finance minister about the bad quality of the raw materials (both tobacco leaves and polvomonte) that arrived at the RTFS, because they were coming in boats without decks, as the finance minister acknowledges in this letter to the superintendent:

I have informed the king about your letter in which you complain about the bad quality that you have noticed in the transportation of tobacco from Cadiz to Seville in boats without decks. (AFTS, leg. 194, letter dated 10 January 1777)

When the king and the finance minister examined this complaint, it was ordered that the transportation of raw materials, polvomonte, and finished products between Cadiz and Seville should be done in boats with decks (AFTS, leg. 194, letter dated 21 February 1777). This meant a higher cost, but the Tobacco Agency managers ensured that both inputs and outputs of the factory were not spoiled during transport.

Case 4: decisions about the sieve machines

The sieve machines were used in the workshops that performed the first and second millings (see Figure 1). Their function was to remove small sticks from the tobacco after the first and second milling processes. Those machines had a significant influence over the quality of the finished products, especially in the case of the second milling (repaso), because after this process the snuff already had the appearance of the finished product. Carmona et al. (1997, 2002) and Gutiérrez (1999) have analysed this case in detail; here we focus on its implications with regard to quality.

In 1776, the superintendent of the factory, Vicente Canasco, carried out several experiments with the objective of improving the efficiency of the sieve machines and thus reducing then cost, without harming the quality of the snuff. The improvement consisted in suspending the sieve from the ceiling of the workshops with ropes. The experiment showed that the Tobacco Agency could save the salary of 114 workers and that even the older workers of the RTFS, who were cheaper for the Tobacco Agency because they earned less, could move the machine "with only one hand" (AHN, leg. 1790, cost statement dated 13 March 1776).

Figure 3 shows the following figures based on an experiment comparing both machines. With the former machine it was necessary to use 26 machines and 272 operators, who earned six reales per day. That amounted to an annual cost (assuming 292 working days) of 476,544 reales. With the new machine, the factory would still use 26 machines but would need only 158 labourers. Therefore, the new machines required 114 fewer operators (at a cost of 276,816 reales). That meant a cost of 199,728 reales less than using the former machines. In addition, there was a saving of 4,650 reales for the cloths used in the sieve machines, as a consequence of a decrease in consumption of 1,550 cloths at a price of three reales, since the new machines involved less friction and therefore less wear and tear on the sieve cloth. Finally, other minor costs dropped by 3,590 reales. That meant that the new sieve machine saved, in total, 207,968 reales. That amount was equivalent to the annual cost of 119 operators (AHN, leg. 1790, cost statement dated 13 March 1776).

The new machines were working on a test basis - measuring the cost and testing the quality - during the months of June, July, and August 1776. The superintendent made a comparison between the cost data for those months of 1775 and the same period in 1776. The resulting lower cost with the new machine was 45,993 reales and nine maravedies. Projected to the whole year 1776, that meant a cost savings of 183,985 reales (AHN, leg. 1790, cost statement dated 4 September 1776).

As the results were good, the finance minister sent Portocarrero, a former operations manager of the RTFS and also the designer of the sieve machines that were to be replaced, to the factory with the objective of inspecting the efficiency of the new machines and the quality of the snuff obtained with them. Portocarrero examined the machines and had several interviews with the superintendent and the workshop foremen between April and May 1776. His report was negative, as the following sentence shows:

I have found that the new sieve machines not only are not profitable as Carrasco said, but they are detrimental for the Crown because of the following reasons: 1st The machines sift very badly ... 2nd The machines produce a lot of waste ... 3rd More work is given to the operators ... (AHN, leg. 1790, report dated 8 May 1776)

This opposition enraged Canasco, who wrote a new report attacking Portocarrero. That meant a quarrel between Portocarrero and the superintendent. Both of them made several experiments with regard to quality, and they even requested the opinion of the foremen of the workshops about the efficiency and product quality of using the new and former machines, with the objective of supporting their positions (AHN, leg. 1970 dated 1 January 1776). The finance minister and even the king himself followed the quarrel closely as copies of the correspondence were submitted to them.

The reports analysed not only cost and quality (AHN, leg. 1970, dated 27 March 1776) but also material waste. Portocarrero said that the new machine produced more dust than the former one (AHN, leg. 1970, dated 10 October 1776); was ergonomically poor because it was larger (AHN, leg. 1970, dated 31 October 1776); and even worsened the nasty smell and temperature of the workshops during certain seasons of the year.

The king sent a commission to study the two models of machines in November 1776:

By order of the king, Lorenzo de Mena [superintendent] has been entrusted to know and prevent the problems existing in the Royal Tobacco Factory of Seville ... and to clarify the case of the new sieve machines. (AHN, leg. 1790, letter dated 21 November 1776)

Finally the king, in an order signed by the finance minister on 20 February 1777, dismissed the use of the new machine because of its bad effects on the quality of the snuff:

The king has known about the whole problem and on the basis of the results of the experiments, which showed the damage caused by the new sieve invented by Carrasco, the king has authorized the use of the former machine invented by Portocarrero. (AHN, leg. 1970, letter dated 20 February 1777)

Case 5: decision to triple production capacity

Portocanero not only inspected the new sieve machines but also designed a new production method with the objective of preventing the raw materials stored in the warehouses from rotting. That meant raising the production capacity in the first-milling stage from around 2m or 2.5m pounds per year to 6m pounds in 1777 (see Figure 1).

Carmona et al. (1997, 2002) have investigated this case. What is remarkable is that this modification worried the Tobacco Agency managers, and even the king himself, about the quality of the finished products. The king followed the modification and even personally analysed the quality of the snuff produced with the new method.

Portocarrero has showed to the king the samples of tobacco he had produced contained in three boxes locked with three keys. (AFTS, leg. 194, letter dated 5 June 1777)

Finally the king approved the modification, although he wanted to be informed about how customers accepted the new snuff in the Spanish kingdoms, and especially in Madrid, because it was an especially large market segment that the king knew personally. That measure is interesting because it is evidence that the managers of the agency took into consideration consumers' opinions for production purposes.

I know that you have started to distribute tobacco elaborated with the new production process by Portocarrero ... I encourage you to report to me about how the customers will accept it, and especially in Madrid. (AFTS, leg. 194, letter dated 18 September 1777)

Case 6: decisions about tobacco stoves

A problem with the accelerated production method developed by Portocarrero was the lack of space to air the tobacco in the fourth phase of the production process (entresuelos: airing) (see Figure 1). The proposed solution was to dry the tobacco artificially through a tobacco-stove system. However, the finance minister warned that, in the years 1728 and 1730, a French man had carried out a similar experiment that was unsuccessful and had been stopped:

In the years 1728 and 1730 there went to this factory a French fellow with the project of drying the tobacco leaves by means of a stove. An experiment was made, and the result was that the tobacco lost its juice and had a disagreeable taste as if toasted. (AFTS, leg. 194, letter dated 10 January 1777)

Thus, the finance minister encouraged the superintendent to conduct the experiment carefully:

The king ordered me to warn you to keep in mind that experiment [the one of 1728 and 1730], and that this time you must avoid making a superficial experiment in order to get reliable data. (AFTS, leg. 194, letter dated 10 January 1777)

The result was that the quality was not good in the long term, and in 1790 the Tobacco Agency suppressed the stoves. In this year, the Tobacco Agency issued a new instruction, the 1790 Instruction, re-establishing the production process that had existed before Portocanero's modification.

Case 7: decisions about cigars

Problems with quality in the cigar workshops were persistent in the 1770s and hampered competition with the products offered by the black market (see Figure 1). Thus, in October 1777, the director of the Tobacco Agency asked the superintendent for reports about the bad quality of the cigars in the years in which Carrasco had been superintendent of the factory. He sent back a negative statement about the quality performance of the RTFS during 1772,1773, and 1775, with a rebuttal from Carrasco.

Look for and send me all the matters that the chief administrator of the Tobacco Agency reported to Carrasco, when he was the superintendent, about the bad quality of the cigars that were produced there. (AFTS, leg. 194, internal report dated 20 October 1777)

Previously the Tobacco Agency managers had ordered the inspection of the cigar workshops because of the poor quality of the cigars, with the objective of avoiding economic losses for the Crown as a result of bad "word of mouth" among customers. This is shown in a letter sent to the superintendent from the finance minister:

The king has seen the bad results of the inspection of the cigars workshops, and he approved a group of measures with the objective of avoiding the existing problems, giving you the order to be vigilant so that neither the public nor the Treasury should suffer the detriment observed. (AFTS, Leg 194, letter dated 21 February 1777)

The consequence of this complaint, besides an order to roll the cigars more tightly, was the development of an experiment, carried out by the trustee of the cigar workshops, with tobacco leaves from Havana and Santo Domingo (Dominican Republic). The experiment involved searching for the best mixture of tobacco leaves to give a finer bouquet to the cigars (AFTS, Leg 194, letter dated 1 May 1777).

6. Discussion and concluding remarks

The cases studied here show the importance of quality in the tobacco business even under monopoly conditions. Attention to quality, which became possible when the problems of the Fdbrica de San Pedro were resolved, was favoured by the Enlightenment spirit aimed at raising efficiency, maximizing state income from tobacco taxation, and, ultimately, fostering public happiness (Sanailh, 1992).

It is true that the monopoly was not absolute, and that one motive for quality control was to create an advantage over smuggled tobacco. Tobacco smuggling was significant in Spain for several reasons: a craggy geography, the poverty of the population, several economic crises and wars, social complicity, the bad quality of the official products, and high prices. Smugglers and defrauders appeared at all social levels: big tradesmen who took advantage of their privileged relationships with power, high clerks and employees of the Tobacco Agency, workers at the RTFS, and finally uncontrolled rural armed gangs and marginal people. Among the measures taken to suppress smuggling was the creation of a specialized police. According to Rodríguez (1994, p.61) this army collective did not achieve the expected efficiency, as it lacked adequate organizational structure and equipment. So the RTFS continued to face significant competition from the black market.

As Figure 1 shows, decisions over quality were taken throughout the production process: from the inputs (the cases of the polvomonte and almagra), through transport (the case of the ships with decks) and transformation on the factory premises (the case of the sieve machines), to the analysis of customers' opinions about the quality of the products (the cases of tripling the production capacity, the tobacco-stove experiment, and the cigar workshop inspections).

The sending of a specialist to Havana despite the cost shows us that quality was a priority for the Tobacco Agency all the way back to the origin point of the raw materials. This measure also facilitated coordination between the Havana factory and the RTFS, which was important to the quality of the finished products, because the tobacco needed to go to Seville just before the production process to avoid being spoiled. Already in the eighteenth century, quality control required a global perspective, from the plantations in Cuba to the customers in Madrid. Other cases of quality concern before the input entered the RTFS were the testing and transportation of almagra and the choice of expensive decked boats to transport tobacco up or down the Guadalquivir.

There were also several quality control points in the RTFS premises with regard to snuff production. Thus, in the second and fifth stages of the production process, first and second milling (monte and repaso), there were technological investments with the objective of increasing the quality of production, as is shown by the design of new sieve machines (Case 4). The same happened in the fourth stage or airing (entresuelos), where the tobacco was dried in stoves, but their installation required a close examination (Case 6).

The interest in quality continued after the tobacco left the factory: after a change in the production process to triple the production capacity for snuff, the Tobacco Agency sought to determine whether the product processed with the new method had the right characteristics for the customers. In the case of the cigar quality problems, even the king wanted to know about the customers' reaction.

One of the most significant quality-related changes required that the operations manager should become ultimately responsible for production processes with regard to product quality, formerly a responsibility of the superintendent. That meant a significant change in the organizational structure, as depicted in Figure 2. The Tobacco Agency even created the post of second production manager to help the senior manager in these matters.

In the case of the new sieve machines, we saw that quality was used as a rationale to settle a personal quarrel resulting from innovation. The Portocarrero-Carrasco feud also had a larger political dimension: this period of Spanish history was characterized by the struggle between conservatives, represented in this case by Portocarrero (a cleric who belonged to a traditional military order), and the Enlightenment reformers, represented by Carrasco (a personal friend of Jovellanos, who was one of the most significant figures of the Spanish Enlightenment). For the purposes of our analysis it is not crucial to know which side was right; what is important is that both sides appealed to quality tests.

The ad hoc measures concerning quality in operations were grounded in two related trends: the reformist spirit of the Spanish Enlightenment, and the administrative reforms of the Bourbon dynasty. Significantly, Canasco and Jovellanos, superintendent and assessor of the factory respectively, were founders of the Economic Society of the Friends of the Country of Seville in 1775, and then reformist compromise between rationalization and the efficiency of the production process was congruent with the society's discourse. In the same vein, as the bureaucratic and obsolete administration of the Hapsburgs gave place to the reformist policies of the Bourbons, the government established royal factories, among which the Tobacco Factory stands out. The objective of these factories was to raise the efficiency of the production process, to increase the royal incomes, and, indirectly, to achieve the happiness of the whole population (Sarrailh, 1992).

Decisions about improving the quality of the finished products and the efficiency of the production process in the factory were concentrated in 1776 and 1777, not only because of poor quality in the previous years but also because of the interest of a particular group of Enlightenment managers. Figure 4 shows the months in which discussions took place concerning these decisions, inferred from the dates of the letters and cost statements.

All these measures were discontinued after the French Revolution (1789). Carlos III had died in 1788, and the Spanish government under his weak son Carlos IV was fearful that the consequences of the Revolution might spread to Spain. Accordingly, the government rolled back all the Enlightenment reformist measures of the previous three decades. At the RTFS, the general instruction of 1790 returned to the principles of the Instruction of 1761, before the reforms of the 1770s (Romero, 1997).