Welfare or politics? The identity of Italian mutual aid societies as revealed by a latent class cluster analysis of their annual reports

Abstract

This article proposes a statistical analysis of the Italian mutual aid societies’ annual report data from 1862, 1873 and 1904, and tests the hypothesis, supported by the mainstream literature, that mutual financial support for the poor is just a facade for organizing the political activity of the working classes. Factor analysis allows the data, signaling these two commitments, to be kept separate. Their time evolution, as outlined by an inter-temporal comparison of a latent class cluster analysis, reveals that political commitment, even if generally prevalent, oscillates as a function of environmental conditions, whereas the mutual financial support retains a solid foundation, regardless. The study demonstrates the potential of unconventional annual report data analysis to reveal the deepest nature of the organizations and contributes to exploring the links between political activism and accounting data in not-for-profit organizations.

Keywords

Annual report Italy latent class cluster analysis mutual aid societies nineteenth century not-for-profit organizations political issues

1. Introduction

Annual report data are, generally, analyzed to define an accurate profile of the financial position and performance of the organizations that produced them. The achievement of this purpose presupposes a clear identification of the type of activity carried out by the organization and its role within the socio-economic context. In contrast, the current study examines the annual report as an instrument capable of rendering the activity of organizations visible to external parties when their nature is uncertain. It applies a reverse process of annual report data analysis, where the real activity of an entire population of organizations represents the dependent variable emerging from their financial position and performance. A historical study of the annual reports of the Italian mutual aid societies (IMAS) in 1862, 1873 and 1904 illustrates this point.

In the second half of the nineteenth century, a large number of mutual aid societies developed in Italy, whose institutional objective was to alleviate widespread poverty through the formation of a fund to help those who had no possibility of working due to illness or disability and provide a livelihood for themselves and their families. In exchange for the payment of an admission fee and periodic contributions, their members had the right to receive help in the form of a subsidy and, sometimes, health care. These associations were often also engaged in other types of activities oriented toward the social and political emancipation of the working classes. However, this secondary commitment was a source of conflict with the public authorities, who tended to consider mutual aid as a facade to conduct political activities that were considered subversive.

This article uses a statistical analysis applied to the annual reports and other non-accounting data of these organizations to elucidate the evolution of the political commitment of the IMAS during the second half of the nineteenth century.

Carnegie and Napier (2012) argue that, among the innovative methods in accounting history, quantitative analysis has been significantly underdeveloped. The article adopts a statistical analysis of large databases containing annual report data and non-accounting information relating to the IMAS for the years 1862, 1873 and 1904, that, over two decades ago, Napier (1989) identified as potentially interesting for accounting historians. The data have been taken from statistical studies published in 1864, 1875 and 1906 by the Italian Ministry of Agriculture, Industry and Commerce, and kept in the Estense University Library of Modena. The data have been subjected to a factor analysis, followed by a latent class cluster analysis, which yields a ranking of the IMAS based on their common characteristics to reconstruct a scenario of the phenomenon for each of the three years surveyed by the Ministry, using Latent GOLD software (Vermunt and Magidson, 2002, 2003a, 2003b, 2005). The resulting three scenarios will be compared, and the time evolution of the activities signaling the institutional identity of the IMAS will be traced.

This article views accounting as a social practice embedded in organizational and social functioning (Hopwood, 1983, 1987; Miller, 1994; Miller and Napier, 1993), which is one of the main thrusts of research into accounting history (Carnegie and Napier, 1996; Miller et al., 1991; Napier, 1989, 2006; Potter, 2005). Compared with previous quantitative studies of accounting history (Chandar and Miranti, 2009; Sivakumar and Waymire, 2003), in which statistical analysis evaluates cause and effect relationships internal to a single organization, the present study uses latent class cluster analysis to reconstruct the evolution of a complete historical scenario. Therefore, the rooting of the quantitative analysis in a deep understanding of the institutional and social environment, recommended by Carnegie and Napier (2012) to avoid a mere application of ‘new’ methods to ‘old’ data, becomes even more crucial; this rooting is the only way to avoid the risk of simple antiquarianism, albeit in a sophisticated way. This risk, according to Napier (1989), is always present when using government reports as research material.

Therefore, to support the investigation, this study considers both primary sources – including government regulations on the IMAS’ accounting and reviews of statutes and annual reports of the most representative IMAS – and secondary sources. These sources offer a clear description of the institutional and social environment and the IMAS’ accounting practices. The primary sources allow us to hypothesize the existence of a decoding system to distinguish the data signaling financial mutual support from those signaling social and political commitment. The study of secondary sources, on the other hand, is central to this article because the research questions originate from the debate in the literature around the true nature of the phenomenon in the late nineteenth century. In fact, mainstream historians highlight the political nature of the IMAS, leading us to investigate the extent to which it characterized the IMAS.

This study has two underlying motivations. First, it contributes to the repositioning of the accounting history literature on the subject of charitable and social institutions and emancipation of the poor (Hollister and Schultz, 2010; Ó hÓgartaigh et al., 2012; Walker, 2004, 2008). It is a response to both encouragement to investigate accounting beyond entities located in the economic base (Hopwood, 1987) and the recent plea for more studies on charitable institutions (Sargiacomo and Gomes, 2011). In particular, this study identifies new opportunities for interdisciplinary exchange between research in accounting history and mainstream history that, in the Italian scenario so far, has found expression limited to state/local government research (Sargiacomo, 2006, 2008, 2009; Sargiacomo et al., 2012). Second, the article demonstrates how quantitative analysis expands the potential of accountability of annual reports, showing how they can express the true nature of organizations with ambiguous institutional profiles. A recent study by Jackson (2012) noted the ability of annual reports to express unusual forms of accountability, which shows that accountability can assume multiple directions and include not only the organization but also the subjects of the surrounding environment; it is a form of accountability created by the same organization, emerging from a qualitative analysis of the annual reports. The form of accountability highlighted by this article is revealed through a quantitative analysis. Some studies have used quantitative data to capture the institutional priorities of an organization. Sargiacomo (2006) used some quantitative data to reconstruct the organization and administration of the feudal municipality of Penne in the late seventeenth century. These studies, however, play a secondary role of completion with respect to the qualitative classification analysis applied to the accounting entries and, in particular, to the expenditure items that are identified as key indicators of the social priorities of the time. The present study, in contrast, considers the qualitative analysis that helps build a decoding system for annual reports as performing a secondary role, supporting the quantitative analysis.

The remainder of the article is organized as follows. Sections 2 and 3 describe the institutional and social environment and the accounting practices of the IMAS, through secondary and primary sources respectively. These sections reveal these organizations’ double identity and identify signals of this in their accounting and non-accounting data. Section 4 presents the hypothesis. Section 5 describes the samples, data, and preliminary analyses. Sections 6 and 7 present the analyzed results for each scenario and general time evolution of the phenomenon, respectively. The final section presents a short summary and concluding comments.

2. IMAS’ double identity: Evidence from the institutional and social environment

After the unification of 1861, the young Italian state seemed full of contradictions: a slow process of industrialization, agricultural backwardness, and imbalances between not only north and south but also from region to region. The IMAS were formed to address the situation of widespread poverty in the country, aiming to help workers in need without offending their human dignity. According to the recurring phrase present in their statutes, their institutional purpose comprised ‘encouraging improvements in the material and moral situation of the praiseworthy working class’. The provision of two areas of improvement gave rise to the dual vocation of the IMAS.

Material improvement of the working class

The IMAS developed initially as voluntary associations of workers where, according to the insurance scheme, economic risks associated with illness, accident, disability and death were shared within the community, with a logic of redistributive solidarity. All members equally contributed, so that even the poorest workers could participate and, in case of need, take advantage of the benefits provided – which usually comprised financial subsidies and, sometimes, provision of medical care. The original purpose of the IMAS, then, was the insurance function; in particular, insurance in case of sickness can be considered their characteristic aid function, although they could differ in other forms of benefits offered. Thus, they mark in Italy the transition from charitable and pietistic-religious forms of aid, where care activities comprised collecting mites and allocating money randomly, to the origins of social security (Cherubini, 1972; Cinelli, 2013; Hernandez, 1972; Gheza-Fabbri, 2000; Gobbi, 1901; Gonetta, 1887; Pessi, 2012; Sapori, 1955). Despite the publication of several textbooks of actuarial science dedicated to them (Gardenghi, 1895; Greco, 1916), the IMAS lacked an effective actuarial system due to absent or imperfect knowledge of the basic rules (or appropriate statistical tables) and/or carelessness in applying them (Cherubini, 1972: 26). Gheza-Fabbri’s study (2000) of the IMAS statutes revealed that they were usually detailed and accurate about admission fees, periodic contributions, and ways of using the available money; however, the rules were always generic and prudent regarding the subsidies to be paid, and their measure was relative to the financial possibilities in the year. The provision of subsidies as a function of financial availability marks the difference between the IMAS and mutual insurance societies (Gobbi, 1901: 54), which adopted a system of guarantees for members because they were liable for the payment of certain sums, and, according to statistical data, technical procedures were established that accurately determined the contributions necessary to enable the association to pay subsidies. The IMAS, however, established contributions according to the economic possibilities of potential members and distributed subsidies according to financial availability. The members could not claim anything in the absence of resources. Historians of the welfare state in Italy identify the IMAS as the precursors of social security because they were aimed at preventing risks instead of waiting for their occurrence and recognized the need to provide aid following the discovery of a state of disease.

Moral improvement of the working class

The literature of the history of the labor movement and trade unions in Italy reveals that, in the late nineteenth century, the IMAS dealt actively with the ‘social issue’ of the living and working conditions of the working class. Although all declared themselves formally apolitical, they gathered the first group of proletariats, claiming better conditions of life and work and posing questions concerning the relationship between the labor movement and political struggle. Therefore, their story intersects with the origin of the trade union and labor movements in the chronicles of mainstream historians (Bonifazi and Salvarani, 1976; Della Perruta et al., 2000; Musso, 1999; Papa, 1976; Perna, 1981; Verna, 2001).

Their first claims are considered the heritage of the suppressed arts and crafts guilds (Cherubini, 1972: 46; Hernandez, 1972: 16; Martuscelli, 1876: 107; Palazzo, 1974: 145; Sapori, 1955: 432) of medieval origin, which were articulated according to the type of activity, had as their purpose the defense of the common interests of their members and protected workers in pre-capitalist Italy. With changes in the production system, they suffered dissolution and then restructuring to become the first Italian workers’ organizations. Between 1770 and 1821, these organizations caused a violent crackdown in the various Italian states when they attempted to regulate work issues and began to exert pressure against the established power. In some cases (Tuscany, 1770; Lombardy, 1772; Sicily, 1821), state laws decreed their abolition. Therefore, the IMAS, having a more professional nature, constituted a place where workers could be protected from repressive regulations on guilds and unions, and promoted team spirit to protect the interests of the working class.

The development of the concept of class struggle passes from the initial stage of the guild to the political and trade union stage (Marucco, 1981; Verna, 2001), which saw commitment to improved working conditions in factories and in the countryside, use of strikes as an instrument of pressure, and provision of support to members in legal disputes with employers. The claims also included the promotion of electoral reform, universal suffrage for the working class, and election of its representatives to the parliament (Verna, 2001). Different political orientations vied for leadership of the IMAS movement (Verna, 2001). The first societies were strongly influenced by Mazzini, and tried to present the movement of the IMAS, which met regularly in regional and national conferences, as a unitary organization able to oppose the monarchy and government. Subsequently, as socialist ideas increasingly developed, the first radical associations, called ‘unions of resistance’, began to emerge. Liberal conservatives tried to relegate the IMAS to pure welfarism and strengthen their ties with the public institutions, while liberal democrats fought for the absolute autonomy of the associations from the government, which continued with awareness and legislative actions aimed at preserving the character of mutual insurance and promoting independence from trade unions, political implications and political organizations.

The government, on one hand, included the IMAS in its repressive actions against the trade unions, and, on the other hand, offered incentives to those who adopted a strict system of mutual insurance to keep them away from a position of making demands (Cherubini, 1972). In fact, the law enacted on 15 April 1886 allowed only the latter access to legal recognition and tax benefits. Even more stringent conditions concerning IMAS’ activity were laid down for obtaining legal personality by royal decree, based on Article 2 of the Civil Code in force during the nineteenth century. Few IMAS demanded legal recognition; moreover, the Ministry’s surveys show that only 17.7 percent of the 6,725 existing associations in 1895 were legally recognized.

3. IMAS’ double identity: Evidence from their accounting practices

The accounting practices of the IMAS in the late nineteenth century attest to their intention to keep their two vocations separate from a financial perspective. The basic principle consisted in the complete separation of the association’s purposes so that each of them corresponded to a special fund with well-defined inputs and outputs (Gardenghi, 1895: 64). The Council of Social Security under the Regulation of 2–3 April 1897, concerning the accounting of the legally recognized IMAS, provided that each category of activity had certain revenues and expenses and a separate fund with a separate accounting system. In no case could the sums belonging to one category of activity could be used for another, even on loan (art. 3). In particular, the fund dedicated to subsidies was inextricably bound to their payment, which could not last for a period exceeding one year. At the end of the year, the fund was closed and its balance was carried forward to the beginning of the following year, but it was still bound to the needs of the members and the society could use it only for the provision of subsidies. The members had to pay a specific contribution for each category of activity and, if the society required a unique contribution, the statute was to determine the way to allocate the revenues to the funds of the various categories. The different categories of subsidies could be represented by independent funds and, if so, the members could contribute to all or some of them.

The regulation also had a strict system for allocating revenue to the funds: honorary members’ contributions, donations and other extraordinary income, when they were not specifically bound, were to constitute a reserve fund to finance the administrative costs and the additional purposes of the association (art. 6). Administrative expenses were, therefore, to be financed by these types of revenue in addition to a special contribution to account separately or as a percentage of ordinary contributions (art. 7). The same article provided that only the income from capital investment, honorary members’ contributions and donations, once the administrative costs were covered, could be used to finance the additional purposes of the association.

This model of accounting was widespread even in the associations that were not legally recognized and represented the vast majority of the IMAS. It was suggested by all the manuals that dictated the accounting rules for the IMAS (Gardenghi, 1895; Greco, 1916) and praised as best practice by the IMAS Federation in its official manual published in 1909. The manuals also illustrated the consequences that the application of these accounting guidelines were to have on annual reports. The latter, in addition to a summary of income and expenses classified by nature, had to explain the breakdown of the entire capital into funds and specify what proportion of each of them was in the hands of the treasurer and how much had been invested. This model of financial management and accounting is summarized in Figure 1.

Figure 1.

The Italian mutual aid societies’ financial management and accounting

Several studies containing reviews of the statutes, regulations and annual reports of the IMAS (Allio, 1980; Gheza-Fabbri, 2000; Papa, 1976; Verna 2001) verify that the inclusion of these guidelines in the statutory rules and regulations was a common occurrence for the IMAS in the late nineteenth century, and their annual reports show various examples of the breakdown of the capital into a part bound to subsidies and a part at the free disposal of the association. Figure 2 shows, for example, a part of the annual report of 1878 of one of the oldest and most influential IMAS, not legally recognized, founded in 1848 in the town of Pinerolo. Moreover, the President of the IMAS Federation, Antonio Maffi, officially expressed his pleasure at the fact that the IMAS largely adopted the accounting model prescribed by law because it served to protect the financial interests of the members (Federazione Italiana delle Società di Mutuo Soccorso, 1909: 89).

Figure 2.

The components of the capital of the mutual aid society of Pinerolo at the end of 1878

4. Hypothesis

The belief that the function of mutual financial support of the IMAS of the late nineteenth century served only as a facade for promoting political goals is widespread in the literature (Cherubini, 1972; Hernandez, 1972; Martuscelli, 1876; Sapori, 1955; Verna, 2001) and is often founded on evidence obtained through analysis of IMAS accounting documents. Cherubini (1972: 25) and Hernandez (1972: 26) argue that they had great difficulty in conducting the mutual insurance task, and that their actuarial devices were inadequate for the purpose. Low wages prevented effective contributions, and the statutes did not contemplate differentiated contributions (compared to wages) against the equality of subsidies. The insufficient number of members prevented the effective compensation of individual risks, making it difficult or impossible to establish appropriate reserves to cope with sudden events. The subsidies were often insufficient compared with needs and, in many cases, were purely symbolic (Verna, 2001). Cherubini (1972) argues that this situation, together with the rule of distributing subsidies within the limits of available funds, should have led them to close their years with no funds for subsidies. However, annual reports surveyed by the Ministry of Agriculture, Industry, and Commerce (1864, 1875, 1880, 1888, 1898 and 1906) showed that a majority of the IMAS possessed considerable liquid capital, signaling additional purposes to be pursued. Moreover, Gheza-Fabbri (2000) observed that, in many cases, the IMAS’ annual reports contradicted their statutes, according to which their main purpose was the provision of subsidies to ordinary members, because expenditures intended for other purposes sometimes exceeded the amount of subsidies paid.

The public authorities perceived the IMAS movement as a political and potentially subversive phenomenon, as evidenced by their surveillance activities and the resulting periodic reports of mayors, prefects, sub-prefects and police (Verna, 2001). The spirit of opposition shown by the IMAS against any interference in their activities by public authorities helped to fuel this perception. Luigi Bodio (founding member, secretary, and then president until 1920 of the Institut International de Statistique) analyzed the reasons for the difficulties encountered in obtaining data concerning the third survey statistics, which began in 1878, as follows:

Many societies refuse to answer questions, not because it requires much time and effort but because of a lively spirit of opposition; because many societies are political rather than oriented to mutual assistance in case of sickness; and because they are troubled by socialist and radical ideas. (Acts of the Central Board of Statistics, 1875. ‘Annals of Statistics’, series 2, 1880, vol. 15: 1)

Relationships between the public authorities and IMAS were always stretched throughout the late nineteenth century. In 1862, the Congress of the IMAS was canceled due to repression led by Prime Minister Urbano Rattazzi. Many societies were dissolved and a fierce press campaign was unleashed, accusing them of disturbing national life. Opposition continued to exist, even in the early decades of the twentieth century when the rise of fascism was marked by the suppression or renewal of control by the fascist party of all movements of a political or trade union character. In particular, repression of the IMAS movement was violent and meticulous (Verna, 2001); many societies succumbed to the violence of the regime and disappeared from the scene. Others survived through acts of obedience to the regime, or by showing themselves harmless. The repressive activity culminated in 1925, when the fascist regime dissolved the IMAS Federation, for its ‘activities openly in conflict with national purposes’.

The hypothesis of the IMAS of the late nineteenth century having an essentially political nature has not yet been proved using empirical evidence; it is, therefore, confined to the field of perceptions. This article contributes to historical studies, providing an empirical verification of this hypothesis with a quantitative analysis of the IMAS annual report data, which compares, at different stages of the movement’s life cycle, the weight of accounting data that indicate activity of mutual financial support with those reporting other activities.

The possibility of quantifying the IMAS, giving priority to their political commitment rather than the financial support of the members, is subject to the possibility of distinguishing within the IMAS’ annual report data a component expressing the mutual financial support from components expressing other activities.

The reasons for this are three-fold; first, the IMAS’ model of financial management and accounting, based on the principle of separation of funds, allows the separation of the two types of activities in terms of accounting data and the identification of political and trade union activities with the management of the reserve fund.

Second, the quality of the data reported in the ministerial statistics allows certain accounting items to be isolated and associated with activities other than the management of subsidies. For example, the statistical report of 1862, the only one which also shows data for the cash in hand at the beginning of the year, explains that this item includes only the part of the reserve fund which is in the hands of the treasurer because it is the only true liquid capital of the IMAS, whereas the liquid part of the other funds belongs to the members (Ministero di Agricoltura, Industria e Commercio, 1864: XXVIII). Moreover, all the statistics (1864: XXVI; 1875: XXI; 1906: 18) clarify that the data concerning cash out-flows for capital investments were excluded from the survey due to the structure of the IMAS’ annual reports. Therefore, it is possible to be reasonably sure that the ‘cash out-flow for other expenses’ item includes only the expenses for the pursuit of non-economic objectives of the IMAS.

Third, some statistical techniques allow a number of ambiguous or ambivalent accounting items to be related to one type of activity rather than another. Considering the IMAS’ model of fund management, it can be reasonably argued that the management of mutual insurance finds its expression in the correlation between the revenue from ordinary members and subsidies paid for sickness, disability, medical care and pensions. However, it is important to note that a percentage, though small, of such revenue could be allocated to the reserve fund.

All other annual report items may, on the other hand, reflect other activities conducted by the IMAS, but not exclusively. The ‘cash in-flow from honorary members’ may be a further sign of potential engagement in additional activities because the principle of fund allocation only allowed the latter to be diverted, but it is characterized by a certain ambiguity or ambivalence since these contributions can also be targeted to cover the payment of subsidies. Honorary members’ contributions, if bound to the payment of subsidies, are in the form of bourgeois philanthropy because the honorary members belonged to the wealthy bourgeoisie and paid regular contributions, without receiving any subsidy, for the sole purpose of integrating the sources of social security. On the other hand, their presence in the IMAS was often aimed at orienting the societies’ claims (Bonifazi and Salvarani, 1976; Cherubini, 1972; Hernandez, 1972; Verna, 2001). They enjoyed presidency of the most important societies and represented them in the national congresses (Verna, 2001). Initially, these leaders were hostile to revolutionary or subversive ideas and refused radical attitudes favoring the class struggle. Over time, however, the moderate trend of the working class declined because of the deteriorating situation of the proletariat (Bonifazi and Salvarani, 1976).

In addition, the ‘other expenses’ were often intended to describe outputs directly explicable in terms of political and trade union activities, such as printing propaganda, legal fees and the costs of organizing political demonstrations (Verna, 2001), but the ‘administrative expenses’ item is generally unclear. It often includes outputs classifiable as ‘other expenses’ (Gardenghi, 1895: 65) and this is the reason why, considering also that these societies operated mainly through the members’ own labor, i.e., free of charge, its amount is often unjustifiably high (Gardenghi, 1895; Gheza-Fabbri, 2000). Therefore if, on the one hand, the accounting items ‘cash out-flow for subsidies’ and ‘cash out-flow for other expenses’ clearly identify two separate kinds of commitment, on the other hand the other types of cash flow may contribute simultaneously to both activities. With the use of statistical techniques, however, it is possible to distinguish the extent to which annual report data, at the aggregate level, converge around the mutual insurance function rather than the non-mutual insurance function. The observation of the relationships of correlation among the variables makes it possible to verify if the ambivalent accounting items tend to be more related, i.e. to move in the same direction, to the payment of subsidies rather than to the ‘other expenses’.

In the light of the foregoing considerations, the general hypothesis relating to a predominantly political nature of the IMAS in the second half of the nineteenth century needs to be developed into two sub-hypotheses:

The internal structure of the IMAS’ annual report data reveals two distinct components, each consisting of revenue and expenditure: one of which is clearly attributable to the function of financial mutual aid while the other relates to other activities; and,

In each scenario, the IMAS, whose component attributable to other activities prevails over that expressing the financial mutual aid, are numerically dominant.

5. Sample selection and data

All the statistics published by the Ministry of Agriculture, Industry and Commerce (1864, 1875, 1880, 1888, 1898 and 1906) demonstrate a significant increase in the number of the IMAS in the late nineteenth century (from 443 units in 1862 to 6,723 in 1898), whereas at the beginning of the twentieth century there was a slight decrease (to 6,535 units). The study is based on three samples selected from the statistics of 1862, 1873 and 1904, taking a census of 443, 1,447 and 6,535 IMAS, respectively. Other editions of the statistical study are excluded because annual report data are absent, or are insufficient to reconstruct the operational profile of the IMAS. After data screening and cleaning, the samples include 375, 1,066 and 5,961 annual reports, respectively, representing 84.65 percent, 73.67 percent and 91.22 percent of the IMAS surveyed in each year. Missing cases, due to societies that did not communicate their accounting data, have been eliminated and outlier values that are well below or above the other scores have been checked and removed due to the sensitivity of the cluster analysis. The annual report data available for each year, and their signaling power in terms of the type of activity of the IMAS, are shown in Table 1. Tables 2 and 3 report summary statistics on annual report data and demographics, respectively.

Table 1.

Available annual report data.

1862	1873	1904
Data signaling mutual insurance activity
Cash in-flow from ordinary members (contributions and admission fees)	Cash in-flow from ordinary members (contributions and admission fees)	Cash in-flow from ordinary members (contributions and admission fees)
Cash out-flow for sickness subsidies	Cash out-flow for sickness subsidies	Cash out-flow for sickness subsidies
Cash out-flow for aid to widows and orphans	Cash out-flow for aid to widows and orphans	Cash out-flow for aid to widows and orphans
Cash out-flow for medicines and medical care	Cash out-flow for medicines and medical care	Cash out-flow for medicines and medical care
Cash out-flow for old-age pensions	Cash out-flow for old-age pensions	Cash out-flow for old-age pensions
Data wholly or partly attributable to other activities
Cash in hand at the beginning of the year (Reserve Fund)	Cash in-flow from capital investments	Cash in-flow from capital investments
Cash in-flow from capital investments	Cash in-flow from honorary members, donations and bequests	Cash in-flow from honorary members, donations and bequests
Cash in-flow from honorary members, donations and bequests	Cash out-flow for administrative expenses	Cash out-flow for administrative expenses
Cash out-flow for administrative expenses	Cash out-flow for other expenses	Cash out-flow for other expenses
Cash out-flow for other expenses

Table 2.

Summary statistics on annual report data (values in lire).

	Cash in-flow from ordinary members (contributions and admission fees)			Cash out-flow for subsidies to the associates and their families			Cash in hand at the end of the year			Cash in-flow from capital investments			Cash in-flow from honorary members, donation and bequests			Cash out-flow for administrative expenses			Cash out-flow for other expenses
	Mean	Median	Std. Dev	Mean	Median	Std. Dev	Mean	Median	Std. Dev	Mean	Median	Std. Dev	Mean	Median	Std. Dev	Mean	Median	Std. Dev	Mean	Median	Std. Dev
1862	1384	729	2203.44	855	326	1761.61	3737	1095	8162.80	244	27	1091.57	393	41.0	951.37	340	128	685.10	203	0.0	916.36
1873	1365	844	1736.57	770	387	1230.86	6106	2169	13947.83	351	71	915.42	258	70.0	571.51	259	122	442.30	229	59.0	533.17
1904	1221	650	1842.22	963	382	1786.92	8818	3000	18044.16	453	84	1221.51	103	0.0	410.31	376	206	521.47	248	68.0	452.55

Table 3.

Summary statistics on demographics.

	Year of foundation			Number of ordinary members			Number of honorary members			Composition			Professional orientation		Geographic distribution			Legal recognition
	Mean	Median	Std. Dev.	Mean	Median	Std. Dev.	Mean	Median	Std. Dev.	Mixed	Men only	Women only	Yes	No	Northern Italy	Central Italy	Southern Italy and Islands	Absent	Present
1862	1851	1857	30.32	225	135	287.01	24	6	42.75	18.9%	77.9%	3.2%	31.7%	68.3%	53.6%	41.3%	5.1%	Data not available
1873	Not available			166	120	152.27	16	7	22.97	22.9%	73.5%	3.6%	16.8%	83.2%	50.9%	36.2%	12.9%	Data not available
1904	1883	1884	15.70	134	90	138.95	Not available			15.8%	80.2%	4.0%	24.7%	75.3%	54.4%	31.3%	14.3%	77.2%	22.8%

The summary statistics highlight the macro trends of the phenomenon over time, which are characterized by the decreasing average size of the IMAS, represented by the mean and median numbers of ordinary members. The progressive reduction of the corresponding standard deviation also shows the increasing tendency of the population toward uniformity around the mean value. An almost similar trend is observed for the number of honorary members. The substantially decreasing trend of the revenues from the two types of members can be explained in light of their numerical reduction.

The subsidies paid and administrative expenses, however, despite a reduction in the average value in 1873, tended to grow in 1904, and there is a very clear tendency of cash-in-hand and other expenses to grow. The latter, in addition, presents a progressively decreasing standard deviation in the tendency of the population to reduce distances from the mean value. Since we do not know the internal details of the administrative expenses, the internal composition of which was rarely specified in the annual reports, it is not possible to say that this demonstrates an increase in activities other than mutual aid. Certainly, however, the IMAS show a growing overall spending power in all types of activities and a greater availability of cash. Against this, there is clearly a growing trend of revenue from capital investments and a decrease of other types of cash in-flow.

Two considerations of synthesis emerge from this picture. First, the IMAS have increased their financial resources over time. Second, they have become more and more independent from the financial point of view by making a partial replacement of the sources of external funding, namely, from the ordinary and honorary members, with internal sources of funding deriving from capital investments.

Regarding demographic characteristics, the prevalence of northern members, men only, and non-professional associations without legal recognition are constants over time, even if the presence of women’s organizations and those established in less-developed areas of the country, the south and islands, marks progressive growth. Moreover, the foundation of a great number of new IMAS in the second half of the century is reflected in the significant drop in the average age of the population.

6. Empirical evidence

The testing of the first sub-hypothesis requires exploration of the underlying structure of the annual report data. The sub-hypothesis is verified if it is possible to ‘reduce’ the mass of report data to two dimensions that represent a double orientation of the IMAS: one toward mutual insurance and one toward activities different from mutual insurance.

To this end, annual report data have been subjected to exploratory factor analysis (EFA) using SPSS version 18. This type of analysis assumes that those annual report data that share similar underlying dimensions should be highly correlated (i.e., they vary together in the same direction), and those data that measure dissimilar dimensions should yield low correlations. Therefore, variables that are correlated with one another, but largely independent of other subsets of variables, are combined into factors (Tabachnik and Fidell, 2007).

The suitability of the data for factor analysis is demonstrated by inspection of the correlation matrix (for each of the three years, this reveals the presence of many coefficients of 0.3 and above), by the Kaiser-Meyer-Olkin value (KMO test), which exceeds the recommended value of 0.6 (Kaiser 1970, 1974), and by Bartlett’s Test of Sphericity (Bartlett, 1954), which reaches statistical significance. The extraction technique of the maximum likelihood factor reveals that the annual report data comprises two underlying dimensions (or factors) that explain more than 70 percent of the variance for each year surveyed (see Table 4 and Appendix 1). All of the most commonly employed techniques, from Kaiser’s criterion to Catell’s scree test (1966), and the most accurate Horn’s parallel analysis (Horn, 1965), converge toward the number of factors to retain.

Table 4.

Exploratory factor analysis (EFA).

	KMO test	Bartlett’s test	Factors	Variables summarized	Factors’ contribution to explanation of variance	Total variance explained
1862	.738	Sig. .000	Mutual Insurance Function (MIF)	Cash in-flow from ordinary membersCash out-flow for financial aid to associates and their familiesCash out-flow for administrative expenses	36.925%	71.041%
1862	.738	Sig. .000	Non-Mutual-Insurance Function (NOMIF)	Cash in-hand at the beginning of the yearCash in-flow from honorary members, donations and bequestsCash in-flow from capital investmentsCash out-flow for other expenses	34.116%	71.041%
1873	.764	Sig. .000	Mutual Insurance Function (MIF)	Cash in-flow from ordinary members;Cash out-flow for financial aid to associates and their familiesCash out-flow for administrative expenses	55.710%	81.383%
1873	.764	Sig. .000	Non-Mutual-Insurance Function (NOMIF)	Cash in-flow from honorary members, donations and bequestsCash in-flow from capital investmentsCash out-flow for other expenses	25.673%	81.383%
1904	.759	Sig. .000	Mutual Insurance Function (MIF)	Cash in-flow from ordinary membersCash out-flow for financial aid to associates and their familiesCash out-flow for administrative expenses	53.259%	78.476%
1904	.759	Sig. .000	Non-Mutual-Insurance Function (NOMIF)	Cash in-flow from honorary members, donations and bequestsCash in-flow from capital investmentsCash out-flow for other expenses	25.217%	78.476%

To aid in the interpretation of factors, orthogonal rotation, with the most commonly used varimax method, was performed. Orthogonal rotation was chosen because it permits uncorrelated factors, which is a prerequisite for an appropriate cluster analysis.

Using the EFA, in addition to confirming the existence of duality in the nature of IMAS activities, overcomes the non-homogeneous data in the three years considered. In each year, there are mutual insurance function (MIF) factors that summarize accounting data unequivocally referable to the activity of mutual financial aid, since it synthesizes the contributions of ordinary members and the subsidies paid to them, and to non-mutual insurance functions (NOMIF) factors.

The administrative expenses are related to the contributions of ordinary members and the subsidies paid to them and this means that, despite the potential ambivalence of these expenses, their predominant content is related to mutual financial aid (e.g., printing membership cards and the associates’ registers, fees for medical certificates). The NOMIF factor is characterized by the item ‘other expenses’ in which expenses for the social and political functions of the IMAS typically flowed. All the cash in-flows different from ordinary members’ contributions are related to this item and, evidently, constitute its main source of funding.

The availability of initial cash, which is only present for 1862, is related to the ‘other expenses’ and confirms, for that year, the orientation to save financial resources for purposes other than subsidies’ management. The existence of this second factor, comprising revenues and expenses that are strongly correlated with each other and weakly related to the MIF factor, demonstrates the existence of the dual nature of the IMAS.

At this point, it is possible to test the second sub-hypothesis, which is confirmed if, in each of the three scenarios, the NOMIF factor prevails over the MIF factor in the majority of the IMAS. It becomes necessary, therefore, to classify the IMAS according to the scores of the MIF and NOMIF factors to quantify, in each year, those that are NOMIF-oriented. Therefore, for each year, the two factors’ standardized scores have been saved in the data files and used to replace the original variables in the cluster analysis. The latent class cluster analyses classify, in each of the three years, the IMAS according to their scores in terms of the two factors, MIF and NOMIF, measuring their orientation toward the mutual and non-mutual insurance functions. It is preferred for the traditional clustering approach to utilize algorithms that group cases that are ‘near’ each other according to some ad hoc definition of ‘distance’ for three reasons. First, it is a probability-based (rather than mathematical) classification, based on the Bayes theorem. Cases are classified into clusters on the basis of membership probabilities estimated directly from the model. Second, variables may be continuous, categorical (nominal or ordinal), or counts, or any combination of these. Third, demographics and other covariates can be used for cluster description. Table 5 shows the role given to the variables, which are distinguished by the software indicators, i.e., dependent variables used to define or measure the latent classes, and covariates, i.e., variables used to describe or predict (rather than to define or measure) the latent classes and, if active, to reduce the classification error. MIF and NOMIF have been assigned the role of indicators to isolate their effects while the other variables have been assigned the role of active covariates. The researcher should indicate the best number of classes but various diagnostics are available to assist in the determination of the number of clusters.

Table 5.

Latent class cluster analysis.

			INDICATORS (standardized values)	ACTIVE COVARIATES (average value for quantitative variables)
			INDICATORS (standardized values)	year of foundation	number of ordinary members	composition		professional orientation		geographical area		legal status
1862	Cluster 1	49.58%	MIF^a = − 0.43NOMIF^b = −0.23	1856	100	women only	4.38%	yes	29.63%	northern Italy	57.27%	data not available
						men only	76.88%	no	70.37%	central Italy	40.30%
						mixed	18.74%			southern Italy and islands	2.43%
	Cluster 2	33.37%	MIF^a = 0.10NOM IF^b = − 0.17	1847	304	women only	2.27%	yes	25.08%	northern Italy	53.81%	data not available
						men only	78.18%	no	74.92%	central Italy	38.72%
						mixed	19.55%			southern Italy and islands	7.47%
	Cluster 3	10.55%	MIF^a = 0.09NOMIF^b = 0.56	1847	320	women only	0.00%	yes	67.52%	northern Italy	41.84%	data not available
						men only	84.01%	no	32.48%	central Italy	48.13%
						mixed	15.99%			southern Italy and islands	10.03%
	Cluster 4	6.50%	MIF^a = 2.63NOMIF^b = 1.73	1847	631	women only	4.14%	yes	40.37%	northern Italy	59.70%	data not available
						men only	73.88%	no	59.63%	central Italy	35.44%
						mixed	21.98%			southern Italy and islands	4.86%
1873	Cluster 1	43.85%	MIF^a = − 0.41NOMIF^b = − 0.63	data not available	92	women only	4.94%	yes	12.37%	northern Italy	44.14%	data not available
						men only	69.64%	no	87.63%	central Italy	43.69%
						mixed	25.42%			southern Italy and islands	12.17%
	Cluster 2	33.73%	MIF^a = 0.09NOMIF^b= − 0.19	data not available	189	women only	2.39%	yes	21.22%	northern Italy	54.81%	data not available
						men only	76.91%	no	78.78%	central Italy	34.83%
						mixed	20.70%			southern Italy and islands	10.36%
	Cluster 3	11.35%	MIF^a = 1.66NOMIF^b = 1.20	data not available	401	women only	1.79%	yes	26.67%	northern Italy	56.45%	data not avai1able
						men only	69.55%	no	73.33%	central Italy	31.74%
						mixed	28.66%			southern Italy and islands	11.81%
	Cluster 4	11.07%	MIF^a = − 0.34NOMIF^b = 0.66	data not available	147	women only	3.54%	yes	10.72%	northern Italy	30.42%	data not avai1able
	Cluster 4	11.07%	MIF^a = − 0.34NOMIF^b = 0.66	data not available	147	men only	82.01%	no	89.28%	central Italy southern Italy and islands	14.51%55.07%	data not avai1able
						mixed	14.45%
1904	Cluster 1	38.19%	MIF^a = − 0.37NOMIF^b = − 0.18	1890	65	women only	3.83%	yes	27.23%	northern Italy	51.81%	not legally recognized 87.43%; recognized by the Court 12.57%; recognized by royal decree 0.00%
						men only	87.58%	no	72.77%	central Italy	27.28%
						mixed	8.59%			southern Italy and islands	20.91%
	Cluster 2	28.15%	MIF^a = 0.23NOMIF^b = 0.03	1879	187	women only	3.63%	yes	23.42%	northern Italy	57.42%	not legally recognized 68.46% recognized by the Court 30.83% recognized by royal decree 0.71%
						men only	74.82%	no	76.58%	central Italy	30.68%
						mixed	21.55%			southern Italy and islands	11.90%
	Cluster 3	25.04%	MIF^a = − 0.33NOMIF^b = − 0.16	1883	84	women only	5.16%	yes	17.05%	northern Italy	53.97%	not legally recognized 77.28% recognized by the Court 22.72 recognized by royal decree 0.00%
						men only	78.72%	no	82.95%	central Italy	38.28%
						mixed	16.12%			southern Italy and islands	7.75%
	Cluster 4	8.63%	MIF^a = 1.87NOMIF^b = 1.59	1871	410	women only	2.68%%	yes	40.21%	northern Italy	57.12%	not legally recognized 60.59% recognized by the Court 35.50% recognized by royal decree 3.91%
						men only	69.25%	no	59.79%	central Italy	30.87%
						mixed	28.07%			southern Italy and islands	12.01%

Notes: ^a mutual insurance function; ^b non-mutual insurance function.

Models from one to four latent classes for each data file have been calculated. The four-classes model has been chosen for each year on the basis of the information criteria (AIC, BIC, CAIC and AIC3). These statistics consider both model fit and parsimony and, when comparing models, the lower the BIC, AIC and CAIC values, the better the model. Moreover, the proportion of classification errors (Class. Err.), which reports the proportion of cases estimated to be misclassified, needs to be kept under control. The closer this value is to zero, the better.

The statistical significance of the set parameters is estimated by Wald statistics, which allows determination of whether all parameters discriminate between the latent classes, and the descriptive power of each parameter is measured by Z statistics, which is used to ensure that the single parameter adequately describes the single class. When standardized bivariate residuals, showing correlations between errors acting on indicators (MIF and NOMIF), show high scores, they are included in the model. These correlations are not explained by the model, which assumes, however, independence of the classes with respect to variables, and their inclusion improves the model fit. For each year, the analysis was repeated several times, with a growing number (up to 1,000) of replications (random sets) and iterations up to the stabilization of cluster size. The results in each scenario show significant Wald statistics (p < 0.05) and Z statistics (z > |1. 96|) for all parameters and acceptable values for the bivariate residuals. Thus, it is possible to define the cluster profiles, as illustrated in Table 5.

The cluster profile identifies for each latent class the average score of the MIF and NOMIF factors, standardized with respect to the average performance of the entire population. This allows for each class to check the intensity of the MIF and NOMIF factors compared with the average performance of the population. Comparing the score of the two factors internally with each cluster, it is possible to identify the areas in which clusters have the best relative performance, and to define its MIF or NOMIF orientations. In addition, the covariates allow us to draw the demographic profile of each cluster and better interpret its nature.

In 1862, the scenario was split into two parts, with 60.13 per cent of the IMAS (Clusters 1 and 3) with NOMIF performance better than the MIF performance compared with the respective mean values of the population, and the remaining percentage (Clusters 2 and 4) in the opposite situation. The active covariates describe, however, a varied situation within the two blocks. The MIF-oriented clusters include the IMAS with the same period of foundation, and differ fundamentally in the number of ordinary members, so that the best performance of Cluster 4 would seem to be explained by the larger average size of the societies. The other covariates show that these clusters are in line with the average characteristics of the population, described in Table 3, even if Cluster 4 stands for a greater presence of professional associations. The NOMIF-oriented clusters, however, are more differentiated. Cluster 1 includes the youngest IMAS, with an average size smaller than that of Cluster 3, and shows a presence of women-only and professional associations, and a geographical distribution in line with the average characteristics of the population. On the other hand, Cluster 3 comprises the oldest IMAS, with a large average size in comparison with the mean and median of the population. Moreover, they are mainly professional men-only associations and with an almost equal distribution between northern and central Italy. This cluster has the typical features of the old guilds of arts and crafts, whose balanced distributions between northern and central Italy are explained by the fact that they were formed mostly by artisans; factory workers were present only in the northern, industrialized part of the country.

The scenario of 1873 shows a greater focus on mutual insurance, with a strong reduction in the number of IMAS favoring the NOMIF factor, which are detectable only in Cluster 4 (11.07%). This cluster is characterized by a contained average size of the societies, by a predominance of men-only societies, and a distribution mainly in the south and in the islands, even if the characteristics are not professional in the vast majority of cases. The other three clusters are MIF-oriented and show performance levels positively correlated to their average size. In terms of composition, they did not differ significantly from the average data of the population. Cluster 3 shows the best overall performance, including the highest percentage of northern IMAS, and that the combination of professional nature and NOMIF orientation disappears.

In the 1904 scenario, NOMIF-oriented IMAS are identifiable in Clusters 1 and 3 for a total of 63.23 percent, and include the youngest and smallest societies, with the highest percentage of societies without legal status. The remaining MIF-oriented clusters show an overall performance directly proportional to the average size and age of the societies. They also show the highest percentage of societies with a legal status. Cluster 4 shows a reduced presence of women-only societies, the largest number of professional associations, and is the only one that collects IMAS recognized by royal decree, which were subject to strict capital requirements.

7. An osmotic relationship with the Italian workers’ movement

The number of NOMIF-oriented IMAS shows a U-shaped evolution over the three scenarios, with a substantial decrease in 1873. This also marks a qualitative change in the phenomenon and leads to distinguishing between NOMIF-oriented IMAS of the old and new generations. The older generation, identified in 1862 in Cluster 3, reflects the heritage of the guilds promoting the interests of specific groups, whereas the new generation that grew back in 1873 was not bound to professional groups and, thus, became an instrument of general social emancipation of the workers. In the contraction phase, the phenomenon appeared mainly in the south and in the islands, traditionally disadvantaged areas of the country. In these areas, the spread of mutual aid societies experienced delay compared with the rest of the country, mainly due to the delay in industrialization, and were used as tools for political and social emancipation of the rural populations (Verna, 2001). The new generation of NOMIF-oriented IMAS then expanded in the early years of the twentieth century, well outside the boundaries of southern Italy and the islands, to become a national phenomenon.

Regarding the MIF-oriented IMAS, their characteristics seem to remain stable in the three scenarios and cluster within each scenario only in terms of average size and performance level. The contraction of 1873, and the subsequent explosion of the new phenomenon, can be interpreted by reading the evolution of the Italian workers’ and trade union movements, to which historians (Bonifazi and Salvarani, 1976; Verna, 2001) bind the fate of the IMAS. The establishment of the First International Workers association in 1864, whose congresses were held in Geneva (1866), Lausanne (1867) and London (1868) and attended by delegates from several IMAS (Bonifazi and Salvarani, 1976: 122), and the peasant struggles of 1868–1869, gave impetus to new forms of association among workers. In particular, the ‘Fasci Operai’ came onto the national scene and, during 1870–1880, linked the experience of the Italian working class to that of the peasant movement. The spread of these organizations, which came together in 1872 in the International Federation of Workers referring to the ideology of the ‘international workers’, had apparently, at least temporarily, emptied the IMAS of vindictive spirit.

The Italian labor movement in the late nineteenth century, however, was subject to diversified growth from region to region, and often even from area to area, because of the different influences of the intellectuals and politicians, and of the specific situations. Where the ‘Fasci Operai’ and the trade unions had not yet arrived, the IMAS became the only means of labor movement available for workers (Bonifazi and Salvarani, 1976: 128). This potential of the IMAS was intentionally exploited by the socialist movement in the early stages of building a trade union as a politically organized movement between 1880 and 1890. The initiatives of the Italian socialist movement were not simply aimed at mobilizing all workers’ organizations, as evidenced by the congress held in Milan in 1891 and attended by the Chambers of Labor and 450 IMAS (Bonifazi and Salvarani, 1976: 146); it was also intended to build a network of IMAS members for the movement. To this end, the Italian Workers’ Party, which was founded in 1882 and evolved into the Italian Socialist Party in 1895, promoted the emergence of new IMAS, which would have a widespread territorial network (Merli, 1972; Verna, 2001). The precise numbers of phenomena are not mentioned, but are defined as general and intense (Merli, 1972: 594) and can largely explain the explosion of NOMIF-oriented IMAS established around 1883 and 1890 in the 1904 scenario. The strength of this phenomenon was probably accentuated by the Catholic response to the clash of class. Simultaneously, Catholics favored the birth of an anti-socialist IMAS right to stem the advance of the socialist movement (Bonifazi and Salvarani, 1976: 167). In addition, in the same years, the Italian welfare state began to emerge, with the establishment of a national fund for voluntary insurance against worker accidents (1883) and a national security fund for disability and old age (1898), giving a first impetus to the retraction of the MIF orientation.

8. Summary and conclusions

This article has attempted to further traditional understandings of the process of annual report analysis, where the quantitative data assume a specific meaning on the basis of the knowledge of the type of activity carried out by the organization. Through statistical tools applied to historic annual report data, the article reversed the process of analysis, deducing the real identity of an entire population of organizations on the basis of their financial position and performance and demonstrating that the annual report data can tell much more than the direction of accountability (Jackson, 2012) or priorities pursued (Sargiacomo, 2006).

This study, specifically, aimed to test the hypothesis, supported by mainstream literature, of a predominantly political nature under the facade of the mutual insurance function of the IMAS in the second half of the nineteenth century. Therefore, by means of an innovative method applied to accounting history, the two factors expressing MIF and NOMIF performance stand out within the IMAS annual report data and provide evidence of their evolution in the period considered. Using an inter-temporal comparison of a latent class cluster analysis based on accounting and non-accounting data for three samples of the IMAS concerning 1862, 1873 and 1904, their MIF and NOMIF performances in three different scenarios were investigated.

The comparison provides evidence of a U-shaped trend of the NOMIF-oriented IMAS, showing that their political and social functions prevail in the first and last scenarios, but not in the second one, at least partially contradicting the initial hypothesis. The contraction phase coincides with the emergence of new forms of worker organizations that initially deprived the IMAS of their political and social content, and a subsequent expansion of the phenomenon confirming the development of a complementary relationship between the IMAS and the labor movement.

The analysis also shows the co-existence of two models of NOMIF-oriented IMAS in 1862. The first reflects the heritage of the guilds of arts and crafts (old generation) while the second is the most open to social and political demands of the new classes of the factory and the peasants (new generation). In the following scenarios, the old generation was destined to disappear, thus adding a full array of NOMIF-oriented IMAS to the history of the Italian workers’ movement. Moreover, despite the prevalence of NOMIF-oriented IMAS in 1862 and 1904, the MIF-oriented associations continued to represent a substantial part (almost 40%) of the entire population. In addition, in each scenario, the NOMIF-oriented IMAS presented scores on two factors, MIF and NOMIF, which are of the same order of magnitude.

From these results, this article draws two general conclusions that contribute to the Italian social and political history of the late nineteenth century. First, mutual insurance cannot be defined as a facade covering political organizations because the orientation toward activities of mutual insurance proves to have strong ‘endurance’, even in the more ‘politicized’ scenarios. It is more appropriate, however, to refer to the real dual personality of the IMAS, with a stable institutional identity comprising both a mutual insurance function and a second, fluid, additional identity that expands and contracts depending on the socio-political context. This identity is easily absorbed by more effective forms of organization that sponsor political and social issues. Second, the literature overestimates the heritage of the ancient guilds in the political nature of the IMAS, since this heritage is of low intensity and limited duration. In fact, since 1862, the political profile of these organizations seems to combine emancipation of the new classes of workers, urban and rural, with the vicissitudes of the labor and socialist movements.

In terms of contributions to international accounting research, this study addresses and intersects two topics of great interest in the recent literature on the subject of accounting history: accounting and charities or similar not-for-profit organizations on one side and accounting and political issues on the other. Some studies have recently addressed the issue of charities/not-for-profit organizations (Baker and Eadsforth, 2011; Baños et al., 2013; Cordery, 2012; Cordery and Baskerville, 2007; Jackson, 2012; Servalli, 2013) and this article adds to the sparse literature on this topic in non-Anglo-Saxon countries.

With regard to the link between accounting and political issues, the current research directions are varied. Some authors have focused on how accounting practices support political power (Jones, 2010; Sargiacomo, 2008; Spence, 2010) and political decisions (Lai et al., 2012); others have explored how accounting systems have changed under the influence of various political ideologies (Ji and Lu, 2013; Rodrigues et al., 2011; Xu et al., 2014) and political shifts (Baker and Rennie, 2012, 2013; Bracci et al., 2010), or the way in which accounting practices have become essential weapons in political struggles (Baños et al., 2013) and the struggle for political legitimacy (Colquhoun, 2013).

One possible intersection between these two research areas has been identified by Servalli (2013) who shows the role of accounting in the interface of power (municipality) and charity in the ‘government of poverty’. The present study identifies a different area of intersection: accounting and political activism in not-for-profit organizations. Even today, not-for-profit organizations, due to their specific guiding principles or special modes or places of action, are often subjects of political importance. The concept of solidarity and mutual support, in particular, inevitably has a political weight, and delimiting the boundary between purposes of mutual help and political activism can be very difficult because of the common basis of values they share. This article, in proposing a method of measuring the political commitment of historical self-help organizations, based on their annual report data, demonstrates the ability of accounting data to express this boundary since they absorb the most hidden nature of organizations well beyond their intentions.

A limitation of this study is that it can quantify the political commitment of the IMAS, compared to their overall activity, but it isn’t able to describe their political role. Political commitment can manifest itself in different ways, ranging from simple propaganda to a vast repertoire of actions which are not easy to exhaustively list. Moreover, the meanings attributed to the concept of political commitment in different periods and places do not always coincide and are strongly influenced by the concrete configuration assumed by political and social conflicts. Due to the lack of detail inside the ‘other expenses’, it is not possible to say if the IMAS were oriented towards simple indoctrination of working-class consciousness rather than direct actions of political claim and how their orientation changed over the course of time. The possible emergence, in future, of this additional information would enable further research to be undertaken. In general terms, promising possibilities for research exist in respect to the attempt to recognize the boundaries of different purposes inside the same organizations, through the lens of accounting.

Footnotes

Appendix 1 Acknowledgements

The author wishes to thank the two anonymous referees for their comments which have helped to materially improve its content. Participants at the seventh Accounting History International Conference also provided valuable comments on an earlier version of the article.

Funding

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

References

Atti della Giunta centrale di statistica (1880) Sessione dell’anno 1875, ‘Annali di statistica’, serie 2° vol. 15. Italian Institute of Statistics Historical Archive (IISHA)

Ministero di Agricoltura, Industria e Commercio (1880) Statistica delle società di mutuo soccorso: anno 1878. Roma: Stamperia reale.

Ministero di Agricoltura, Industria e Commercio (1888) Statistica delle società di mutuo soccorso e delle istituzioni cooperative annesse alle medesime: anno 1885. Roma: Tipografia Metastasio.

Ministero di Agricoltura, Industria e Commercio (1898) Elenco delle società di mutuo soccorso. Roma: Tipografia della casa editrice italiana

Ministero di Agricoltura, Industria e Commercio (1864) Società di Mutuo Soccorso. Anno 1862. Torino: Tipografia Letteraria.

Ministero di Agricoltura, Industria e Commercio (1875) Statistica delle società di mutuo soccorso: 1873. Rome: Regia Tipografia.

Ministero di Agricoltura, Industria e Commercio (1906) Le società di mutuo soccorso in Italia al 31 dicembre 1904: studio statistico. Rome: Tipografia Nazionale di G. Bertero.

Ministry of Economy and Finance Historical Archive (MEFHA) Atti del Consiglio della Previdenza – Sessione del 1897 (Roma: Ministero di Agricoltura, Industria e Commercio. Divisione credito e previdenza).

Pinerolo Mutual Aid Society Historical Archive (PMASHA) ‘Pinerolo Mutual Aid Society’ Collection, ‘Accounting and Finance’ Series, Bundle 4.

10.

State Central Archive (SCA) Legge 15 Aprile 1886, n. 3818 – ‘Costituzione legale delle società di mutuo soccorso’.

11.

Allio

(1980) Società di mutuo soccorso in Piemonte, 1850–1880: attività economica, gestione amministrativa, ambiente sociale. Torino: Deputazione Subalpina di Storia Patria.

12.

Baker

Eadsforth

(2011) Agency reversal and the steward’s lot when discharge exceeds charge: English archival evidence, 1739–1890. Accounting History 16(1): 87–109.

13.

Baker

Rennie

(2012) Accounting for a nation’s beginnings: Challenges arising from the formation of the Dominion of Canada. Accounting History 17(3–4): 415–435.

14.

Baker

Rennie

(2013) An institutional perspective on the development of Canada’s first public accounts. Accounting History 18(1): 31–50.

15.

Baños Sánchez-Matamoros

López-Manjón

Carrasco-Fenech

Funnell

(2013) Symbolic capital, accounting and caciques in local political life: The charity of Mr Rafael Tenorio (1909–20). Accounting History 18(4): 491–506.

16.

Bartlett

(1954) A note on the multiplying factors for various chi square approximations. Journal of the Royal Statistical Society 16(series B): 296–298.

17.

Bonifazi

Salvarani

(1976) Dalla parte dei lavoratori, storia del movimento sindacale italiano. Milano: FrancoAngeli.

18.

Bracci

Maran

Vagnoni

(2010) Saint Anna’s Hospital in Ferrara, Italy: Accounting and organizational change during the devolution. Accounting History 15(4): 463–504.

19.

Carnegie

Napier

(1996) Critical and interpretive histories: Insights into accounting’s present and future through its past. Accounting, Auditing & Accountability Journal 9(3): 7–39.

20.

Carnegie

Napier

(2012) Accounting’s past, present and future: The unifying power of history. Accounting, Auditing & Accountability Journal 25(2): 328–369.

21.

Catell

(1966) The scree test for number of factors. Multivariate Behavioral Research 1(2): 245–276.

22.

Chandar

Miranti

(2009) Integrating accounting and statistics: Forecasting, budgeting and production planning at the American Telephone and Telegraph Company during the 1920s. Accounting and Business Research 39(4): 373–395.

23.

Cherubini

(1972) Storia della previdenza sociale in Italia. Roma: Editori Riuniti.

24.

Cinelli

(2013) Diritto della previdenza sociale. Torino: Giappichelli.

25.

Colquhoun

(2013) Political and organizational legitimacy of public sector auditing in New Zealand local government. Accounting History 18(4): 473–489.

26.

Cordery

(2012) Funding social services: An historical analysis of responsibility for citizens’ welfare in New Zealand. Accounting History 17(3–4): 463–480.

27.

Cordery

Baskerville

(2007) Charity financial reporting regulation: A comparative study of the UK and New Zealand. Accounting History 12(1): 7–27.

28.

Della Perruta

Misiani

Pepe

(2000) Il sindacalismo federale nella storia d’Italia. Milano: Francoangeli.

29.

Federazione Italiana delle Società di Mutuo Soccorso (1909) Manuale per società di mutuo soccorso. Milano: Tipografia Cooperativa Varesina.

30.

Gardenghi

(1895) Manuale tecnico per le società di mutuo soccorso. Milano: Hoepli.

31.

Gheza-Fabbri

(2000) Solidarismo in Italia fra XIX e XX secolo. Le società di mutuo soccorso e la casse rurali. Torino: Giappichelli.

32.

Gobbi

(1901) Le società di mutuo soccorso. Milano: Società Editrice Libraria.

33.

Gonetta

(1887) Le società di mutuo soccorso e cooperative in Europae specialmente in Italia, loro origine, scopo, utilità. Torino: Loecher.

34.

Greco

(1916) Le società di mutuo soccorso. Torino: Unione Tipografico-Editrice Torinese.

35.

Hernandez

(1972) Lezioni di storia della previdenza sociale. Padova: Cedam.

36.

Hollister

Schultz

(2010) Slavery and emancipation in rural New York: Evidence from nineteenth-century accounting records. Accounting History 15(3): 371–405.

37.

Hopwood

(1983) On trying to study accounting in the contexts in which it operates. Accounting, Organizations and Society 8(2–3): 287–305.

38.

Hopwood

(1987) The archaeology of accounting systems. Accounting, Organizations and Society 12(3): 207–234.

39.

Horn

(1965) A rationale and test for the number of factors in factor analysis. Psychometrika 30(2): 179–185.

40.

Jackson

(2012) ‘The collector will call’: Controlling philanthropy through the annual reports of the Royal Infirmary of Edinburgh, 1837–1856. Accounting History Review 22(1): 47–72.

41.

(2013) The evolution of bookkeeping methods in China: A Darwinist analysis of developments during the twentieth-century. Accounting History 18(3): 317–341.

42.

Jones

(2010) Sources of power and infrastructural conditions in medieval governmental accounting. Accounting, Organizations and Society 35(1): 81–94.

43.

Kaiser

(1970) A second generation Little Jiffy. Psychometrika 35(4): 401–415.

44.

Kaiser

(1974) An index of factorial simplicity. Psychometrika 39(1): 31–36.

45.

Lai

Leoni

Stacchezzini

(2012) Governmentality rationales and calculative devices: The rejection of a seventeenth-century territorial barter proposed by the King of Spain. Accounting History 17(3–4): 369–392.

46.

Martuscelli

(1876) Le società di mutuo soccorso. Firenze: Le Monnier.

47.

Marucco

(1981) Mutualismo e sistema politico, il caso italiano (1862–1904). Milano: FrancoAngeli.

48.

Merli

(1972) Proletariato di fabbrica e capitalismo industriale: il caso italiano 1880–1900. Scandicci (Firenze): La Nuova Italia Editrice.

49.

Miller

(1994) Accounting as a social and institutional practice: An introduction. In: Hopwood

Miller

(eds) Accounting as Social and Institutional Practice. Cambridge: University Press, pp.1–39.

50.

Miller

Napier

(1993) Genealogies of calculation. Accounting, Organizations and Society 18(7–8): 631–647.

51.

Miller

Hopper

Laughlin

(1991) The new accounting history: An introduction. Accounting, Organizations and Society 16(5–6): 395–403.

52.

Musso

(1999) Tra fabbrica e società: mondi operai nell’Italia del Novecento. Milano: Feltrinelli.

53.

Napier

(1989) Research directions in accounting history. British Accounting Review 21(3): 237–254.

54.

Napier

(2006) Accounts of change: 30 years of historical accounting research. Accounting, Organizations and Society 31(4): 445–507.

55.

Ó hÓgartaigh

Tyson

(2012) ‘Irish property should pay for Irish poverty’: Accounting for the poor in pre-famine Ireland. Accounting History Review 22(3): 227–248.

56.

Palazzo

(1974) Le società operaie di mutuo soccorso. Mandria: Lacaita editore.

57.

Papa

(1976) Origini delle società operaie in Piemonte. Milano: Giuffrè.

58.

Perna

(1981) Breve storia del sindacato: dalle società di mutuo soccorso al patto federativo. Bari: De Donato.

59.

Pessi

(2012) Lezioni di diritto della previdenza sociale. Padova: Cedam.

60.

Potter

(2005) Accounting as a social and institutional practice: Perspectives to enrich our understanding of accounting change. Abacus 41(3): 265–289.

61.

Rodrigues

Schmidt

dos Santos

Dutra Fonseca

(2011) A research note on accounting in Brazil in the context of political, economic and social transformations, 1860–1964. Accounting History 16(1): 111–123.

62.

Sapori

(1955) I precedenti della previdenza sociale. In: Sapori

Studi di Storia Economica. Firenze: Sansoni, pp.427–441.

63.

Sargiacomo

(2006) Using accounting records to enhance an understanding of a seventeenth-century Italian feudal community: The case of the commune of Penne (1664–90). Accounting History 11(4): 475–502.

64.

Sargiacomo

(2008) Accounting and the ‘art of government’: Margaret of Austria in Abruzzo (1539–1586). European Accounting Review 17(4): 667–695.

65.

Sargiacomo

(2009) Accounting for ‘the good administration of justice’: The Farnese state of Abruzzo in the sixteenth century. Accounting History 14(3): 235–267.

66.

Sargiacomo

Gomes

(2011) Accounting and accountability in local government: Contributions from accounting history research. Accounting History 16(3): 253–290.

67.

Sargiacomo

Servalli

Carnegie

(2012) Accounting for killing: Accountability for death. Accounting History 17(3–4): 393–413.

68.

Servalli

(2013) The interface of power and charity in the government of poor: A case from the Italian context in the sixteenth–seventeenth centuries. Accounting, Auditing & Accountability Journal 26(8): 1306–1341.

69.

Sivakumar

Waymire

(2003) Enforceable accounting rules and income measurement by early 20th century railroads. Journal of Accounting Research 41(2): 397–432.

70.

Spence

(2010) Accounting for the dissolution of a nation state: Scotland and the Treaty of Union. Accounting, Organizations and Society 35(3): 377–392.

71.

Tabachnik

Fidell

(2007) Using Multivariate Statistics, 5th edn. Boston: Pearson Education.

72.

Vermunt

Magidson

(2002) Latent class cluster analysis. In: Hagenaars

McCutcheon

(eds) Applied Latent Class Analysis. Cambridge: University Press, pp.89–106.

73.

Vermunt

Magidson

(2003a) Latent class models for classification. Computational Statistics and Data Analysis 41(3): 531–537.

74.

Vermunt

Magidson

(2003b) Latent GOLD Choice 3.0 User’s Guide. Belmont, MA: Statistical Innovations Inc.

75.

Vermunt

Magidson

(2005) Latent GOLD 4.0 User’s Guide. Belmont, MA: Statistical Innovations Inc.

76.

Verna

(2001) Società operaie di mutuo soccorso. Pescara: Ires Abruzzo Edizioni.

77.

Walker

(2004) Expense, social and moral control: Accounting and the administration of the old poor law in England and Wales. Journal of Accounting and Public Policy 23(2): 85–127.

78.

Walker

(2008) Accounting, paper shadows and the stigmatised poor. Accounting, Organizations and Society 33 (4–5): 453–487.

79.

Cortese

Zhang

(2014) Ideology diffusion and the role of accounting: A Gramscian approach to understanding China’s transition from 1949 to 1957. Accounting History 19(4): 434–451.