Contractual relationships and accounting change: The case of Aguera wholesalers, 1770-1835

Introduction

Among studies investigating both the process and the outcome of accounting change, some have addressed internal issues, such as the influence of technology (Chandler, 1962), replacement of financial managers (Martinez Ruiz, 1988; Boyns & Edwards, 1996), tightening of control mechanisms (Carmona, Cespedes & Gomez, 1997), changes in factory layout (Carmona, Ezzamel & Gutierrez, 1997), financial crisis (McKendrick, 1970; Hopwood, 1987), and company growth (Johnson, 1983). Other studies have analysed the potential impact of environmental factors such as intellectual and social movements (Bhimani, 1994) and competition (Edwards & Newell, 1991; Fleischman & Parker, 1991). Still other researchers have investigated how internal and external factors intertwine in fostering accounting change (Carnegie, 1993, 1997; McWatters, 1995). Most of the literature mentioned above focuses on businesses which are identified as being large. Furthermore, in some of these instances the number of employees is given (for example, Carnegie, 1997); a feature which can be used as a reference regarding an organisation's size (Damanpour, 1992). Utilising the number of employees as criteria, we can strengthen the argument that many historical studies in accounting are mainly concerned with large-scale businesses. As Fleischman and Parker (1991) pointed out, we have much to learn about how accounting systems historically functioned in small to medium-sized organisations. Moreover, in the literature on accounting history, the few existing studies of such organisations have focused on family-owned firms in mining, farming, and dock activities (Napier, 1991; McLean 1997) - that is, generally not commercial trade.

This paper reports the case of Almacenes Aguera (Aguera Wholesalers), a small to medium-sized, family-owned company group that operated in the grocery wholesaling industry, both retail and wholesale, in the Spanish province of Cadiz. The antecedent of the company group was a single retail shop established in 1740 by Juan de Aguera, a trader from the port city of Santander in northern Spain. In 1770 his son and successor, Jose de Aguera, along with two other partners from Santander, Jose Gonzalez de la Sierra and Jose Bolivar Ydoeta, founded a wholesaling business, through a private agreeement between the partners. Although this association was never formally constituted, it would eventually form the nucleus of a larger company group,1 supplying merchandise to shops inside as well as outside the group. The employees, comprising only five or six people at the central business, were also natives of Santander, and the company provided them with accommodation, food, and clothing. Half of the initial capital was supplied by Jose de Aguera and 25 per cent by each of the other two partners. The ownership and control structure of the main wholesaling business is particularly interesting for the purposes of this paper. As the three partners of the firm and their descendants were normally resident in Santander, some of them periodically had to travel the 1,000 kilometres to Cadiz in order to monitor the main business. The monitoring process consisted of (1) direct supervision of the business, normally, by one male member from one of the three families, who periodically alternated in performing this task; and (2) bi-annual meetings, during which accounts were rendered, dividends were distributed, and managers were appointed for the subsequent two-year period. In other words, there was some separation between management and ownership, a separation that increased after 1822. In 1835, the endpoint for this study, ownership became concentrated in a single family.

This study examines the impact of contractual relations within Aguera Wholesalers on changes in the firm's accounting system. We build on the theory of "relational contracts", which focuses on contracts affected by relations that are not purely economic (for example, family bonds between agents). Although this theory provides considerable explanatory power in this case, we also show some limitations of the theory in addressing the relationships under consideration. This paper is anticipated to interest accounting historians for a number of reasons. First, although accounting history research has focused on phenomena related to relational contracts, such as interpersonal trust, to explain past practices (for example, McWatters, 1995), this study is one of the first to address distrust as well as trust. This framework may illuminate complex contracts like those involving family bonds, which affect many businesses. Second, investigation of small to medium-sized companies is rare in accounting history, especially for firms operating in retail and wholesale distribution. Lastly, as Hernandez-Esteve (1995) asserts, investigations addressing the accounting practices of nineteenth-century Spanish companies are non-existent, and we believe that this study may shed some light on the implementation of double-entry bookkeeping in firms as a result of a change in commercial regulation during this period.

The paper is organised as follows. First, we introduce the foundations of relational contract theory. Second, we describe the archival sources that underpin this study. Third, we outline the economic, financial, and social contexts of the company group, giving particular emphasis to the changes in its ownership structure. We then describe the changes in the accounting system over time. Finally, in the last section we explore the underlying reasons for those changes and draw conclusions.

Relational contracts and accounting

Agency theory constitutes a valuable framework for accounting research (Baiman, 1982, 1990). The principal-agent framework has provided testable propositions to address issues, such as the demand for accounting reports, accounting choice, accounting regulation, and the stock market reaction to mandated accounting changes (Hunt & Hogler, 1990). In spite of these achievements, agency theory is still surrounded by controversy (Wiseman & Gomez-Mejia, 1998). The notion of the agent, which is crucial to the theory, is also subject to close scrutiny; the assumptions of rationality and economic motivation cast serious doubts on the generalisability and applicability of the theory. Moreover, most research has focused on particular principal-agent relationships (Jensen & Meckling, 1976), and has neglected many other interesting relationships affecting the firm (for example, peer relationships). More recent theorists have argued that contracts are not just direct consequences of firms' economic objectives, but arise from a number of closely interconnected bilateral relations (Wiseman & Gomez-Mejia, 1998).

Whereas the economic perspective of agency theory identifies optimal contracts that maximise the utility of both principal and agent, relational contract theory embraces a number of non-economic phenomena such as family relationships and friendships between agents (MacNeil, 1978). It views the agency contract as a relationship that is inherently informal, experiential, ambiguous, implicit, based on trust, and involving factors different from purely utilitarian concerns (Gomez-Mejia & Wiseman, 1997). Relational contract theorists argue that the interaction between agent and principal creates an exchange value as well as an intangible utility that benefits the relationship. Further, the parties' experience in the contractual relationship creates mutual expectations for future interactions. Drawing on those expectations, the parties envision the relationship in an open-ended manner, and this, in turn, increases their commitment to the relationship. Finally, personal ties and the indefinite time horizon of the relational contract promote commitment and resolution of disagreements arising from low performance. Relational contract studies contend that the principal will try to delay the establishment of specific performance indicators and to avoid assessing the agent's performance exclusively upon such a basis.

Relational contracts are closely related to relational trust (Bigley & Pearce, 1998). Relational trust is based on a psychological contract that regulates parties' behaviours and creates a common bond as well as a set of mutual expectations (Rousseau, 1989). Trust is defined as the intention to be vulnerable and reliable to the other party and is regarded as an alternative to pure control mechanisms. Principal-agent contracts are influenced by relational trust in several respects. First, the relationship embraces a long-term horizon that assumes agents will retain their jobs despite poor short-term performance (Rousseau et al., 1998). Second, relational trust "forestalls the monitoring and evaluating of a partner's behaviour" (Bigley & Pearce, 1998, p.413). This implies that there are few, if any, sanctions in place to curb opportunistic behaviour by the agent. Third, relational trust leads to more positive evaluations of the other party's behaviour and motivations, even despite negative information. Moreover, relationships are not uni-dimensional, and the interaction frequency, duration, and diversity of challenges that relationship partners encounter may produce a condition in which elements of trust and distrust persist. Under these circumstances, relational contract theory predicts that parties will take steps such as establishing monitoring mechanisms to bind those facets of the relationship engendering distrust (Lewicki et al., 1998).

Agent-principal family ties represent a special case of relational contracts. Relational trust in the context of a family-owned business derives from kinship and often direct blood ties between agent and principal. In this context, management and ownership may not be completely separate; nevertheless, the agency problem may still arise. In a classic paper on ownership structure, McEachern (1975) found that agency problems are as evident under conditions of high shareholder dispersion ("management controlled" firms) as they are under conditions of high executive ownership ("owner managed" firms). Relational trust becomes a substitute for formal monitoring systems, and performance evaluation is based on subjective assessments (for example, negative results are justified by appealing to factors beyond the agent's control). Disciplinary measures are seldom enacted, and this brings about stability of management. In short, the literature on relational contracts assumes that the agent has a stronger influence on the relationship, and this causes the agent to bear less risk and be subject to softer control mechanisms than those prescribed in principal-agent theory. This situation poses an extra incentive for the agent to become opportunistic. As far as accounting systems are concerned, the following characteristics may be predicted in relational contracts:

* control is reduced, and this reduction influences disclosure as well as the frequency of accounting information demands;

* decisions are not continuously legitimated, so there is less use of accounting;

* long-term management appointments imply an expected low rate of change in accounting systems;

* accounting systems are expected to play a low-profile role in decision making.

Primary archival sources

The archival materials used in this study have been drawn from the Historical Archives of the Gonzalez de la Sierra Family (Historical Archives of the Province of Cadiz, Spain). Table 1 lists those of the company group's files that we consider relevant to the period covered in this study: not only accounts but other valuable qualitative materials like wills, company deeds, private correspondence, and books containing copies of all business correspondence. Taken together, these materials are well preserved and provide a comprehensive depiction of the company's activities. They allow us to ascertain both the technical aspects of accounting systems in use and how the systems affected other organisational processes.

The company and its context: 1770-1835

When Jose de Aguera and his two partners set up their wholesaling business in 1770, they were taking advantage of favourable economic conditions created in 17)7, when Cadiz was granted a monopoly of Spanish commerce with Latin America (Garcaa-Baquero, 1988). Thereafter, the city enjoyed a "Golden Century" that attracted traders from other parts of Spain as well as from other European countries (Garcaa-Baquero, 1972). Population increased by 86 per cent over the course of the eighteenth century and, between 1778 and 1792, exports increased especially rapidly. More specifically, from 1783 the end of the war with England brought about a tremendous increase in Maritime trade in Cadiz. This lasted up to 1786 when all the markets, which had dried up during this conflict, again became saturated (Solis, 1978). Generally favourable conditions lasted until 1810. From this point onwards, the siege of the city by the French and a prolonged fall in prices coinciding with the start of the independence movement contributed to the ruin of trade in Cadiz. At the same time, the high duties imposed on the Port of Cadiz diverted trade to the nearby port of Gibraltar. Between 1815 and 1825, the city's population diminished by 33 per cent (Ramos Santana, 1992). By 1824, and coinciding with the end of Spanish sovereignty over continental America, 36 per cent of the trading houses in Cadiz - 235 establishments - were bankrupt (GarcaaBaquero, 1972). Even though Aguera Wholesalers was not directly involved in the colonial trade, its second and, especially, third-generation managers faced extremely difficult business conditions.

Jose de Aguera died without issue in 1771, one year after the establishment of the main wholesaling business. he named his six sisters as his heirs, but appointed his two partners in the organisation, both trusted friends, as executors of the estate and administrators of the company group for an indefinite period of time. His last Will and testament attempted to anticipate possible conflicts among the heirs. From Jose de Aguera's point of view, tranquillity and harmony between his heirs would bring the "benefits of peace" that he considered necessary to "the progression of the business" (A.H.G.S., Jose de Aguera's Will, Cadiz 20 December 1768, Case 795, Series: Private documentation, Notarial and Legal). He accordingly forbade any interference on the part of his heirs with the administration by his executors:

I hereby forbid my inheritors to oblige the executors to make any judicial inventories or other similar actions, since my intention is to save my successors from unnecessary costs and my executors from nuisance [...] and, if any of my heirs do not accept this condition, it is my wish that they should be excluded from the inheritance.

[...] It is my wish that my heirs contact my executors who will inform them of the current state of the wholesaling business, as well as that of the shops, without under any pretext making judgements of the management. What they are told by my executors should always be accepted. (A.H.G.S., Jose de Aguera's Will, Cadiz 20 December 1768, Case 795, Series: Private documentation, Notarial and Legal).

However, Aguera's nephews would be allowed to join the firm's management if they complied with certain requirements:

If any of my nephews shows managerial skills to run the business, I entitle the executors to include them in the administration of the wholesaling business and assign them a fair salary. It is also my Will that if my nephews prove their managerial competence, they may take over the position of the deceased executors and the subsequent control and management of the wholesaling business (A.H.G.S., Jose de Aguera's Will, Cadiz 20 December 1768, Case 795, Series: Private documentation, Notarial and Legal).

Accordingly, among Aguera's five nephews, the partners appointed the two who were evidently most interested in business2 to join the organisation: Pedro Aguera in 1771 and Benito Gonzalez Tanago in 1772. Thus, between 1770 and 1822, management was evenly carried out, in an alternating manner, by the three families which shared the organisation's ownership. However, at any one time, there was normally only one person in charge (see in Figure 1 the families to which the individuals who alternated as managers during these years belonged). In practice, this implied that at the end of each accounting period, the general manager appointed for the following period moved from Santander to Cadiz along with one or more representatives of the other families, the latter taking over reporting responsibilities.

The last two decades of the eighteenth century witnessed the firm's highest earnings (for example, profits for 1785-1786 were 217,884 Reales', see Figure 2). Furthermore, the number of shops almost doubled compared with those existing during the early years, whereas no new wholesale business was opened (see in Table 2 the number of shops in 1770 and 1785). However, wholesaling still provided 60 per cent of the partners' total dividends.3

In 1820, after the death of Jose Bolivar Ydoeta and in accord with his last Will, Juan Antonio Campuzano, a Santander-born trader who was the father of Ydoeta's son's wife and the tutor of the Ydoeta grandsons, was named to represent the Ydoeta heirs in Aguera Wholesalers (Ydoeta's son having died before he could be named an executor). In 1822, Jose (III) and Francisco Gonzalez de la Sierra, both grandsons of the founder of that name, but not yet owners, took over the management of the company group. Soon thereafter, during the commercially disastrous period of 1824-1826, Francisco Gonzalez de la Sierra became the sole administrator.

The recession caused the closure of a number of the Aguera establishments. As Table 2 shows, the number of Aguera shops in Cadiz fell from seven to three between 1820 and 1828. Figure 2 shows the steady decline in profits. However, relational trust contributed to a subjective assessment of managerial performance: the owners (principal) supported their agent by attributing the low performance to factors over which the agent had no control (Lewicki et al., 1998). As Campuzano, the representative of the Bolivar Ydoeta family, wrote on 27 September 1825 to the manager of the Aguera Wholesalers, Francisco Gonzalez de la Sierra:

Trade is in serious decline. It is not surprising thus that profits from the shops are stagnating, and this trend will remain until things recover and return to normal levels. Man can only do what he is capable of, the rest must be in the hands of God (A.H.G.S., Santander 27th September 1825, Case 820, Series: National Correspondence).

In 1828, the Aguera family withdrew of the main wholesaling business and it went into liquidation.4 However, the business continued operating under a different ownership structure. Formally, Campuzano, in Santander, and Jose (III) Gonzalez de la Sierra, in Cadiz, took over management. In fact, Campuzano became the owners' spokesman, whereas Jose (III) Gonzalez de la Sierra, and to a lesser extent his brother Francisco, became the company group's managers. In the course of the Aguera withdrawal, the capitalisation of the central wholesaling business was doubled: the Ydoeta and Gonzalez de la Sierra families increased their shares from one-quarter to one-third each, while more than doubling their actual investments (from 75,000 to 200,000 Reales), and the Gonzalez de la Sierra brothers, without waiting for the approval of the remaining two families of heirs, assigned themselves a one-third ownership at the price of 200,000 Reales. Campuzano, on behalf of the Santander owners, acquiesced in this arrangement. From then on, the company group concentrated on wholesaling5 (see Table 2).

The 1830s witnessed a crucial conflict between the Santander-based owners, represented by Campuzano,6 and the company group's management, the Gonza7lez de la Sierra brothers. Jose (III) Gonzalez de la Sierra attributed the conflict to jealousy on the part of the Santander-based partners because of his and his brother's purchase of capital shares in the main wholesaling business:

You will surely be aware that I have made many enemies among the Agueras for the liquidation of the old company, and I am also conscious that many of the De la Sierra family are not in a good mood with me, because they are all jealous of my interest in the company [...] (A.H.G.S., letter from Jose Gonzalez de la Sierra to Campuzano, Cadiz, 15 January 1830, Case 821, Series: National Correspondence).

Indeed, Campuzano was suspicious about the effective payment-in-capital of the Gonzalez de la Sierras' stake. Wariness and scepticism, as expressions of the owner's distrust, have important potential effects on accounting and other monitoring mechanisms (Lewicki et al., 1998). In this case, the owners' spokesman also claimed that the reporting system provided him with irrelevant and unreliable information. Jose (III) Gonzalez de la Sierra replied:

You suggest that you have not seen that my brother and myself have put up the money corresponding to our one-third share in the capital of the company. This mistrust encompasses all operations and upsets me, as far as it questions my perfectly honourable behaviour [...] (A.H.G.S., letter from Jose Gonzalez de la Sierra to Campuzano, Cadiz, 15 January 1830, Case 821, Series: National Correspondence).

One month later, the successors of Jose (I) Gonzalez de la Sierra and Bolivar Ydoeta raised some concerns about the firm's management. First, both families were unhappy with the company's new name and demanded the inclusion of their own family names. Although the brothers argued that "business names seek brevity and dynamism", in the end they agreed to name the company "Ydoeta, Sierra y Cia" (A.H.G.S., letter from Jose Gonzalez de la Sierra to his brother, Francisco, 12 March 1830, Case 821, Series: National Correspondence). Second, the Santander-based owners complained about the draft of the firm's bylaws, which established that the new ownership structure should be maintained for ten years and that withdrawals from capital should require the prior consent of other owners. Third, the owners objected to the proposed compensation of managers, suggesting, at first, a fully variable system, and afterwards, a fixed one. With respect to the former proposal, Jose Gonzalez de la Sierra argued:

It would be unfair to leave the managers without remuneration, or to give them only a small amount if the company's yields were minimal or non-existent. After all, they have dedicated all their efforts to the company [...], a situation that would put the managers in the dilemma of deciding their own salary levels (A.H.G.S., letter from Jose Gonzalez de la Sierra to Campuzano, Cadiz, 15 January 1830, Case 821, Series: National Correspondence).

These issues demanded a rapid decision because the new deed had to be registered in the Registre Publico y General de Comercio (General Commerce Public Register), as required by the 1829 Code of Commerce, which established uniform commercial regulations for all Spanish territory.

In May 1830, Francisco Gonzalez de la Sierra met with the remaining owners to discuss the conditions for the formalization of the new company. He and Campuzano agreed on the clauses to include in the company's bylaws, or at least such agreement seemed to be implied in a letter sent by Campuzano: "Please, find attached an authorization to proceed with the approval of the company's bylaws. This authorization is subject to the introduction of some clauses that were presented to me by your brother, and signed by your father and other interested parties [...]" (A.H.G.S., letter from Campuzano to Jose Gonzalez de la Sierra, Santander, 14 May 1830, Case 821, Series: National Correspondence). Nevertheless, the Gonzalez de la Sierra brothers never registered the company bylaws.

The 1830s witnessed serious financial problems for Ydoeta, Sierra y Cia., which, in 1831, incurred losses for the first time in the firm's existence (see Figure 2). In a purely principal-agent relationship, the owners could have blamed the managers for these results. However, the additional influence of relational trust contributed to a positive evaluation of the agent's performance (Lewicki et al., 1998). Managers and owners alike attributed the firm's losses to external factors: "when one has no control over events, the only thing one can do is to accept what God offers" (A.H.G.S., letter from Francisco Perez de la Sierra, former manager of the company group, to Francisco Gonzalez de la Sierra; Santander, 12 December 1831, Case 821, Series: National Correspondence). There was thus a short period of peace between managers and owners (1831-1833), as the correspondence between the two sides demonstrates.

In 1833, Jose (III) Gonzalez de la Sierra took over control of the businesses, an event that was followed by quarrels between management and ownership. In 1834. Jose (III) forced out Campuzano, arguing that the Gonzalez de la Sierra family should continue as the owners of the main wholesaling business and citing what were essentially family reasons: his father was still alive, and some other family members were already taking an active part in management. As a result, in 1835. the Bolivar Ydoeta family sold its 33 per cent share of capital to Jose (III) Gonzalez de la Sierra for 200,000 Reales. An additional payment of 3,835 Reales was made as a contribution to fixed assets (Book 183, Series: Official Books of Accounts). The Gonzalez de la Sierra family together with the Gonzalez de la Sierra brothers were now in possession of all the capital.

The accounting system

Accounting regulation experienced significant developments during the eighteenth century and the early-nineteenth century. The Ordinance of Commerce of Bilbao, issued in 1737, mentioned the double-entry bookkeeping method as an option. The Tribunal of Commerce of Bilbao7 was the first to issue ordinances regulating the trading activities of merchants, and other Spanish Tribunals of Commerce often adopted its rules (Hernandez-Esteve, 1992, 1996). In 1806, the Tribunal of Commerce of Sanlucar de Barrameda (Cadiz) enacted an ordinance that largely resembled Bilbao's standards. Not until the Code of Commerce of 1829, however, were all Spanish companies required to implement the double-entry bookkeeping method - and even then, the requirement was virtually toothless.

Our description of accounting practices embraces four periods. The first period (1770-1821) witnessed fairly active control by the remaining living founders.8 The second period (1822-1828) encompassed the entry of the Gonzalez de la Sierra brothers into management and the withdrawal of the Aguera successors from the main wholesaling business. The third period (1829-1830) saw the establishment of a new ownership structure, a concentration on wholesaling, and the beginning of conflict between management and owners. The fourth period (1831-1835) brought the implementation of significant accounting changes and the concentration of ownership within the Gonzalez de la Sierra family. The accounting books in use during these periods are shown in Appendix 1.

1. Stability and expansion of the businesses (1770-1821)

The accounting system was designed to render accounts to partners about both profits and the distribution of profits. To this end, aggregated statements were produced on an irregular basis. For example, whereas the first 22 years of the company group's existence witnessed annual/bi-annual reporting, after 1793 this frequency slumped to every seven/eight years. The existing system also attempted to implement a set of internal controls over the heirs' wealth, assets, and liabilities.

Beginning in 1770, these objectives were fulfilled by the following accounting books: The First Book, Book of Inventories, Small Cashbook, Provisional Daybook, Fifth Book, and Book of the Shops. In 1808, when the company group opened its second wholesaling business, two more books were introduced, concerning the purchase of and payment for goods at the central wholesaling business and the balance sheets. Since correspondence throughout this period does not indicate any complaints by the owners concerning profit information, we contend that the accounting system satisfied the owners' information needs.

2. Changes in management (1822-1828)

In 1822, Jose (II) Gonzalez de la Sierra retired to Santander and delegated his son Jose (III) as representative of the family's interests in Cadiz. As a result, Jose (III) and his brother Francisco managed the main wholesaling business during a period of low profit for the company (see Figure 2). Given that the brothers did not hold a share in the capital, the separation between ownership and management increased - and with it, reporting to ownership and innovations in accounting (see Figure 1). These changes aimed at providing more detailed information about the company group's establishments at outlying locations. At the central wholesaling business, Francisco kept individual accounts for each customer (including other Aguera establishments) in a number of Sales Books; debts were then entered into the Second Book. This information also shed light on taxes due (at that time, taxes were calculated on the basis of the shipment of goods in and out of towns). By the time these improvements were introduced, however, the Book of Purchases ceased to be used.

3. Constitution of the new company and concentration on wholesaling (1829-1830)

After 1828, the reconstitution of the company, the concentration on wholesaling, and the quarrels between partners about the overall management of the company group brought further changes in accounting. In 1829, Campuzano made clear to Jose (III) Gonzalez de la Sierra his expectations about how the accounting system should work:9

Since the books of the extinguished company are not applicable to the new one, it will be necessary to keep new books: one to record purchases, either on cash or on credit; another to copy exactly word by word the balance sheets of the central wholesaling business, adding the amount to be debited or credited to the partners in their respective accounts, noting the book and page. Another book should keep accounts of the establishments located outside Cadiz and another for accounts of customers located in Cadiz and supplied by the firm. Another book to keep the accounts with partners, and credit or debit them as appropriate; and another to keep account of the income and the cost produced by the houses; and finally, another to copy all commercial correspondence. All of them, except the last one, must be lined and numbered by page, as well as having their purposes written on the front page (A.H.G.S., Santander, 13 January 1829, Case 820, Series: National Correspondence).

Thus in 1829, a catalogue of the new books was made: Books 1, 2, and 3 replaced the former First and Second Books; Book 4 fulfilled the same function as the previous Small Cashbook; and Book 5 monitored the entry of goods into the warehouse. This last book was a novelty, since the company had stopped recording the intake of merchandise by 1824. In general, however, there were very few accounting innovations, and Campuzano's suggestions were ignored.

In January 1830, Campuzano's son was appointed as controller of the new company. At the same time, Jose (III) showed a genuine interest in adapting the company group's accounting system to the Code of Commerce of 1829, which required firms to implement the double-entry bookkeeping method. Nevertheless, the Aguera Wholesalers' accounting methods remained the same, as can be seen in a letter from Campuzano received by Francisco Gonzalez de la Sierra soon after taking over his brother's post. Campuzano demanded information about his principal's interest in the businesses and complained about the quality of accounting information, specifically requesting a copy of "the last balance sheet completed at the central wholesaling business. I also expect to receive timely information about charges made to the successors of Mr Bolivar Ydoeta" (A.H.G.S., Santander, 21 September 1830, Case 821, Series: National Correspondence). Francisco Gonzalez de la Sierra answered as follows:

Although all your comments are very relevant, [...] the accounts as kept until now have been clear enough. For the moment, I do not have in mind sending you the requested accounts, at least, not as quickly as you demand. If I did, would the rest of the shareholders not follow your example, for they too have equal interests? Of course they would! The information already sent to you is sufficient to satisfy your needs (A.H.G.S., Cadiz, 26 October 1830, Case 821, Series: National Correspondence).

4. Accounting changes and ownership-management conflicts (1831-1835)

During the early 1830s the quarrels related to the constitution of the new company were creating a certain level of distrust. Several accounting changes were introduced in response to some of the suggestions made by Campuzano in 1829. Thus, between 1833 and 1834, the following books were opened: Invoices Book, Draft Daybook of Sales, Book of Shipments to Customers outside Cadiz, Journal, Cashbook, Cadiz Customs, and Correspondence Book.

Accounting books now tracked the flow of merchandise, monitored every consignment, and specified debts and cash collected. The procedure was as follows: the acquisition of a consignment was first noted in "Book 5" and some more detailed information then was recorded in the Invoices Book. If the merchandise was sold in Cadiz, the sale was recorded both in the Invoices Book (as a sale) and in the Draft Daybook of Sales (as a revenue). Otherwise, the transaction was recorded in the Invoices Book (as a sale), in the Book of Shipments to Customers outside Cadiz (as a revenue), and in the Sales Books (which tracked transactions with each client). Lastly, the receivables and debts were recorded in the Journal and the charges and payments in the Cash Book.

In 1834, Campuzano was still complaining about the reporting system (A.H.G.S., letters written in Santander, 2 September and 14 October 1834, Case 822, Series: National Correspondence); the Gonzalez de la Sierra brothers never fully met his requests.

Analysis

During the early years of Aguera Wholesalers, both the careful provisions contained in Jose de Aguera's Will and the economic situation of the Bay of Cadiz provided a supportive context for the company group's development. The Will required inheritors to fully trust the founding partners, and to renounce any control over them, evident from the prohibition to judge the founders' management and the obligation to accept their decisions. The early accounting system of the company group measured the yields of each establishment and reported profit information to the company's owners. Accounts were provided up promptly until 1793 and the profitability of the main wholesaling business was high. The chief accounting innovation during the first fifty years was the implementation of the Book of Purchases, in 1808, immediately after the opening of the company's second wholesaling business.

After 1800, ownership control over management became even softer, with less concern about reporting to shareholders and the subsequent distribution of profits (accounts were rendered only thrice in twenty years). This lack of concern, we contend, may be explained by the withdrawal of Jose Bolivar Ydoeta from active involvement in the businesses (see Figure 1). Despite this lack of information, there is no evidence of any complaint made by the inheritors. The favourable economic situation ensured the continuation of the owners' trust in management. This increasing confidence, enforced by Jose de Aguera's Will and by family ties, had contributed from the beginning to the development of relational contracts between managers and owners that forestalled any problematisation of accountability to the owners (Bigley & Pearce, 1998).

Thus, during the first fifty years of company group's existence - a period characterised by economic growth, expansion of the business and high returns on capital - a working relationship of trust was generated between principal and agent that fits well with the postulates of relational contract theory. Jose de Aguera's Will also contributed to this process. The owners relaxed the control mechanisms, since they were satisfied with the company group's performance (McLean, 1997) and trusted fully in management (Lewicki et al., 1998). Consequently, the accounting system did not undergo any substantial modifications and was kept fairly simple.

More substantive accounting innovations were made in the 1820s. In a context of economic stagnation and changes in management control, Francisco Gonzalez de la Sierra took over the management of the company group. A number of new accounting books were opened in that year, in order to track business by geographical areas and comply with the fiscal system of the time which charged local taxes only on the goods sold to be consumed in the same town. Reporting frequency returned to the former standard of every two years. As previous studies have demonstrated, in Aguera Wholesalers the presence of a new manager brought about some accounting innovations (Martinez Ruiz, 1988; Boyns & Edwards, 1996). Moreover, we argue, the greater separation between ownership and management (see Figure 1) lessened trust on the part of those owners who were not relatives of the managers. As a consequence, the managers felt psychologically obliged to render more accounts than had their predecessors (Rousseau, 1989).

In the subsequent economic crisis of 1824, the trading activities of the area collapsed. The average annual profits of the main wholesaling business decreased from 71,460 Reales between 1807 and 1822 to 14,100 Reales between 1822 and 1829. Seven-eighths of the retail establishments went bankrupt, forcing the company group to close down most of them while establishing three wholesaling businesses. Yet, despite the managers' anxieties, their family ties and their long history of interactions10 fostered the owners' trust in them (Rousseau et al., 1998), so that the negative results were justified by appealing to factors beyond the managers' control (Lewicki et al., 1998).

The shift towards wholesaling and the increase in the number of wholesaling businesses required an increased monitoring of purchased goods. A new book was opened in 1829 to account for this; receipts were chronologically recorded and the suppliers were explicitly mentioned. In short, the change in business strategy demanded new information. Greater geographical distances between the establishments provided additional motivation for these accounting changes.

The preceding analysis shows that an increase in reporting to shareholders and the earliest changes in the accounting system coincided with a decrease in the profits of the main wholesaling business, slumps in dividends, and the death of the remaining founder and executor, among other factors. These findings have two implications. First, they permit us to concur with researchers (for example, McKendrick, 1970 and Hopwood, 1987) who contend that financial crises propel changes in the accounting process. However, whereas in manufacturing firms, such as Wedgwood, accounting figures created new means for understanding an organisation's activities, in a commercial business, such as Aguera Wholesalers, accounting changes were oriented to external activities given that internal activities were uncomplicated. Second, as previous historical studies have shown (for example, Napier, 1991), the management of an organisation can be affected by clauses in the founder's Will. Specifically, as long as at least one of the executors of Jose de Aguera's Will was alive, the accounting changes were less frequent than after both of the executors had died. During his lifetime, Jose Bolivar Ydoeta, who was backed by Jose de Aguera's Will, could have impeded succeeding managers from reforming the accounting system designed by himself.

At the end of the 1820s, the Agueras gave up their interest in the main wholesaling business. As a result, the Gonzalez de la Sierra brothers acquired a large share of the capital and took over the control of the company group. This new ownership structure increased the internal conflicts between the families, for several reasons: the shift away from the practice of rotating management, the absence of the other families' representatives from the management group, the reduction in absolute terms of profits distributed to the owners,11 and suspicions about the brothers' ability to pay cash for their share in the capital. Moreover, the managers enjoyed a high degree of autonomy due to their high level of ownership in the firm's capital and the organisation's weak control systems. As a result, the managers no longer benefited from the moral authority of the founding partners.

Distrust imposed tighter monitoring on management. First, Campuzano sought closer monitoring of management by suggesting specific changes in the accounting system. His demands produced only one change, and it was mainly window dressing. Some real modifications and improvements (for example, the accounting book for controlling the purchasing of goods, and the creation of an independent book to record the heirs' dividends) were introduced when the books of the old company group were closed and new books were opened to serve the same functions. Second, Campuzano's son was sent to Cadiz to take over as controller - evidently, in order to exercise a more direct control over management. Lastly, whereas the two managers wanted the "Ydoeta, Sierra y Cia" company's deeds to allow incentives tied to management performance, the owners wanted to continue the familiar practice of fixed salaries:

If we keep to such a clear path [the existing fixed salary for management], which is the method that has regularly been followed in all establishments of this type, why should we depart from it and follow a different one? One which naturally will create its own difficulties, and certainly be more misleading? In my opinion, the existing method is the one which should be continued (A.H.G.S., letter from Campuzano to Jose (III) Gonzalez de la Sierra, Cadiz 30 March 1830, Case 821, Series: National Correspondence).

From the perspective of agency theory, this behaviour apparently contradicts the owners' self interests, as the proposal to link remuneration to results was also of potential benefit to them. However, the existing level of distrust precluded agreement.

The 1829 Code of Commerce required companies to institute double-entry bookkeeping. Initially, the managers proclaimed themselves anxious to comply with the Code:

I mentioned in one of my former letters that the matter of the bylaws was not urgent [...], but I had no knowledge at that time of the new Code of Commerce. This delay is now causing a great deal of trouble with the Consulate here (A.H.G.S., letter from Jose Gonzalez de la Sierra to Juan Antonio Campuzano, Cadiz, 16 April 1830, Case 821, Series: National Correspondence).

However, their position changed in the face of the owners' demands for more, and more relevant, information. The difficulty of the double-entry bookkeeping method was converted into a justification for not meeting those demands. Accounting information became a source of conflict between ownership and management. The owners cited their need to keep track of their own current account. The managers cited the difficulty of putting a new system into practice in a small to medium-sized firm (lack of qualified personnel, waste of resources, and time required to train personnel in the new accounting techniques) and their own satisfaction with the existing system. As the risk of fines was negligible, the Code hardly introduced incentives for adopting the new method.12 In short, the brothers were reluctant to satisfy demands for information from the other owners.

The developments of the 1830s, we contend, evince patterns of trust and distrust typical of a mature relationship. The accounting system originally based on a trusting relationship was no longer operative because changes in the ownership structure and the company group's management had increased distrust and limited trust to certain aspects of the relationship. On the one hand, the owners wanted to monitor management more strictly through accounting information related to those matters where greater distrust existed (Lewicki et al., 1998). On the other, the managers delayed both the implementation of the double-entry bookkeeping system and changes in the existing accounting system in order not to be accountable to the owners. Despite the managers' share in the capital, their objectives did not align with those of the owners (McEachern, 1975). The owners did not have sufficient information concerning the actions of the managers and suspected that the latter's behaviour was opportunistic. Therefore, the owners attempted to control management actions through monitoring mechanisms related to the day-to-day work and decisions - for example, by establishing a mutual vigilance between managers (Oviatt, 1988), and by demanding more accounting information. Interestingly, however, they made no attempt to align interests between the different parties by linking compensation to performance (Tosi et al., 1997). In contrast, they sought a flat compensation scheme because of distrust, information asymmetry, and lack of control mechanisms over management. We infer that their intention was to prevent the agents from manipulating the profits (Lewicki et al., 1998).

In this case, long managerial tenure weakened owners' trust, as is evident in their preference for a fixed salary. When confidence weakens, agency problems appear. For example, a tightening of controls, more information, and timely reporting are demanded; agency costs are increased (the appearance of the controller); and new changes in the accounting system are demanded.

The losses in 1831 (see Figure 2), we contend, led to changes in the accounting system from 1833 to 1835: tight control over inventory levels, monthly reports to enforce strict customer monitoring, tracking operations by functional areas, and control of merchandise purchased from overseas suppliers.

Conclusion

The preceding analysis provides some rationales for the notion that accounting changes may be encouraged by the interplay of environmental and internal factors (see McWatters, 1995). In particular, this study has shown that social factors condition individuals' economic conduct. Principal and agent do not act as atomistic individuals isolated from their social context; rather, they are embedded in a network of specific, ongoing social relationships (Granovetter, 1985). For precisely this reason, this study should be read cautiously. We doubt whether eighteenth and nineteenth-century concepts of opportunism and morality were the same as those of today. Some authors (Cowton and O'Shaughnessy, 1991; Mills, 1993) have indicated that solutions regarded as suitable today because of their implications for organisational efficiency would be inexplicable in past eras, and that individual self-interest has been put forward as a basic norm of social life only since the second half of the nineteenth century. Up until then, relationships possessed a strong moral component, being defined in terms of custom, duty, and accountability. However, other authors (for example, Noreen, 1988) state that when principal and agent comply with a set of moral values, reinforced by sanctions and by religious or other precepts, they do not do so without any self-interest. In this study, we observe clear economic interests hidden behind the accounting information. In sum, we agree with Mills (1993) that the most a historian of accounting can do is to incorporate other possible explanations that may help to clarify the results, without trying to theorise in other areas of the social sciences. Furthermore, given the lack of information about accounting practices in companies of this era in Spain, we have little knowledge of the context with respect to individual behaviour, agency relationships, and accounting systems. It would be well worthwhile to study the adoption of (or resistance to) the double-entry system by other similar-sized nineteenth-century Spanish companies - especially given that throughout the 1830s, the single-entry method, together with the double-entry method, continued to be taught at the Business School of Cadiz (Capelo et al., 1996).

Other directions for future research related to contractual relationships emerge from this study. First, ownership concentration in a single extended family from 1835 on encourages further research from the relational contracts perspective. Focusing on the accounting and private information available would illuminate the relationship between accounting changes and contractual relationships. Second, another logical extension of the study would be to elucidate our statement that, in family companies where trust is based on kinship, decisions are not continuously legitimated. This would imply a more limited use of accounting information for decision making. By studying the accounting information available, we should be able to analyse the purposes for which it was used, and whether its use could have conditioned or been conditioned by personal interests and opportunistic behaviour. Lastly, the interrelationships among types of contracts, compensation, and accounting systems have not been explored to any material extent in the business history literature. Making use of available archives, of the genre examined in this study, makes such investigations possible, thus permitting a stronger focus to potentially emerge of such interrelationships.

Acknowledgements: The authors gratefully acknowledge the financial support of Project 97-13.58 corresponding to the Programa Sectorial de Promocoon del Conocimiento (Ministry of Education and Culture, Spain). They are also grateful to three anonymous referees and the editor, Garry Carnegie, for insightfulness and assistance.