The word organization indicates a concept mainly referred to a social structure having three specific aspects:

·        a plurality of components - such as humans and materials or both -;

·        quite steady relationships

·        a goal

The word "organization” has a secondary meaning both in the English and Italian languages that is organization as an activity. Namely we use the expression “organization of a conference" or "lack of organization" meaning that in order to reach an objective it is necessary to collect resources and properly allocate them. As a result, it is evident that the three basic aspects characterizing the term “organization” - plurality of components, steady relationships and intended goal - are also common to the secondary meaning of this word.

Trade unions are organizations and we can see the three, above- mentioned aspects in them. However, being representative organizations, trade unions also present additional, more specific aspects. They are associations and their members choose to join together and build that kind of organization to defend their interests, strengthen their position in society and improve their working and living conditions in general. These organizations give members a voice that would otherwise be too faint or lacking at all. These features and goals are common to every association having a representative function and also affect the nature of these organizations, their structure, their internal relationships, as well as the sources and distribution of resources. Studies regarding representative organizations, such as trade unions, highlight two structural characteristics:

The next two sections offer a detailed explanations of these characteristics.

Loosely coupled systems

As above said, organizations have "steady relations". This does not mean that they are necessarily rigid, but that we can expect them to be - at least to some extent - formal, never casual but always lasting in time. To what extent and in which manner relationships within organizations are "steady" depends on the different kind of organization, and consequently on their different goals, components, forms and strategies. The binding strength of internal relationships within an organization is a fundamental quality as it is used by researchers to classify organizations and identify a specific group, namely loosely coupled organizational systems. Trade unions falls within this group.

The “ideal organization” is commonly regarded as perfect machinery, for instance as a clock with every single cog working together to turn out its final product: the exact time. Translating this metaphor into the social scene the ideal organizations are armies and private enterprises. Therefore, if we observe an organization which functions differently we may be inclined to think that it is working badly. However, this might not be the case.

Paradoxically, it is possible to use a question to explain how organizations can function in different ways. According to your experience, whenever a clear decision is made within a representative organization, such as a trade union, can an exact fulfilment of the requirements ensuing from such decision be expected?

It depends. May be yes, maybe only in part, maybe completely not, maybe later. All this uncertainty about results surely disturbs; someone will be angry and will desire to work in a different organization. But why this uncertainty? It is not due to malfunction but rather to the complexity and variability of the representative function. Members of representative organizations are both owners and clients, suppliers and recipients of these organizations.

As a result, this kind of organizations does not have just one priority, nor can they use one single principle to make decisions and overcome conflicts. In an imaginary organization there is a clear cause-effect relationship between elements. Trade unions, instead, do not follow this criterion and the relationship between two elements does not always depend on one variable, as it is the case with private enterprises, which have efficiency as a must and whose top priority is making profit. In our organizations efficiency is important, but we could decide to lose money and/or time in order to achieve an immaterial and uncertain result. For instance, activists, whose work is a relevant resource, have no guarantee in terms of results and reward, and are free in their decision to start and stop their contributions.

The economic rationality of cost and benefit analysis doesn't always give answers that can explain the actors' behaviour in social and political scenarios. We find two opposite views among researchers. Basically, on the one hand Olson maintains that there is an incentive to participate in any collective action if expected benefits outweigh expected costs. On the other hand, Hirschmann states that being active in social and political participation is satisfactory in itself, regardless of concrete results. Possible results if any will bring additional satisfaction. Both attitudes can be found in individuals and the prevalence of one view over the other varies according to different contexts, mainstreaming culture and economic situation. However, representative organizations seem to prefer the second approach and regard their presence and participation as achievements. Achieving concrete results is indeed important but is not always the sole priority for these organizations. Furthermore, there can be structures within the same organization acting with different logics at the same moment. For these reasons the concept of the "loosely coupled system" is useful to understand how a trade union works. In this kind of organization there is a changeable combination of variables whose effect one can add or cancel other ones in different contexts and at different times. This notion was introduced into organizational studies by Karl Weick and it defines a system where each element has a certain grade of autonomy and can vary the influence received by or exercised on the other components of the system. Loosely coupled systems are more complex than the simplified ideal organization characterized by a well defines, strong, centralized rule and by a machine-like perfection. However, complexity does not mean malfunctioning.

Table 1. Qualities affecting relationships between elements within each system.

rigid/strong tied systems	loosely coupled systems
continuous constant relevant direct immediate	sudden changeable minor indirect subsequent

Table 1 shows the qualities affecting relationships between elements within each system. As far as time is concerned, the relationship between elements is continuous in rigid systems, whereas it may be absent, sudden or discontinuous in loosely coupled ones due to the autonomy of their single elements. As for the grade of influence, it is constant in rigid systems and changeable in the other ones. When comparing internal elements of both systems with external factors, the influence of internal elements in the organization functioning are more relevant, direct and immediate in rigid systems than in loosely coupled ones, which are much more sensitive to the environment where they are located and operate.

We generally think of a pyramid when we imagine the shape of an organization. The organization map can be represented in different ways, but, in any case, we have less people at the top than at grassroot level. As we climb the organizational ladder powers become less specific and wider ranging and the quantity and quality of controlled resources is increased. Every step of the pyramid is supposed to confer more power. There is a hierarchy given that an individual covering a super ordinate role in the organization has the power to make binding decisions for those having a subordinate role. Issuing fiats is an important mechanism to solve internal conflicts and, when accepted, it solves them quickly. This is very important when time is scarce. On the other hand, by excluding any discussion, this approach results in missing any potentially useful contribution coming from subordinates.

Table 2 compares hierarchical interdependency in rigid tied organizational systems and in loosely coupled ones. In the latter systems it looks like weak, sometimes almost absent. It could seem that while the latter is based on consent, the former relies on coercion. However this is not the case. Consent exists in both systems although it is expressed differently. Whereas in rigid systems consent leads to follow fixed rules and procedures, in loosely coupled systems consent is aimed at achieving goals and leaves more autonomy with regards to the means and methods used to reach them. As a consequence of increased autonomy in loosely coupled systems the use of sanctions is prevented. Consequently, hierarchy is rather seeming than real, works only in single parts of the organization, without an overall control on it. In most cases decisions are issued by collective bodies rather than by individuals, and the power of individuals comes from collective decisions. The reduced power of applying sanctions makes binding decisions into exhortations, which are not effectively compulsory.

rigid/strong tied systems	loosely coupled systems
real general/overall compulsory individual	seeming partial impelling collective

At this point a loose coupling organizational architecture seems to be dangerous and ineffective. This is partly true but it also brings two kinds of benefit. The first is that the autonomy of internal parts makes them more adaptive to their environment, a very important quality for a representative organization that only autonomy allows any single part of an organization to achieve. The second benefit is that autonomy reduces the risk that a change made within one element will create unanticipated changes within other elements. Limiting interconnections can help isolate problems when things go wrong and simplify testing, maintenance and troubleshooting procedures. On the contrary, the limit is that it is impossible, or at least very difficult and slow, to implement a decision that could improve procedures, results and efficiency in every part of the organization.

The picture below shows the different flows of resources in a trade union, which is a loosely coupled organization, and in an enterprise, which is a much more rigid tied system. In trade unions resources come from members who give their association fees and their mobilization to a local and/or sector structure, where a part is kept and another goes to upper levels up to the top level. In the enterprise the flow of resources follows an opposite direction: the owners gives them to the top level and from there they arrive to the lowest one. As a result a trade union, whose members are free to decide whether to associate and remain or leave, is condemned to democracy. Being a representative organization means representing the members' preferences, otherwise sooner or later, these members will leave.

Ruling loosely organizational systems, such as trade unions, is surely more difficult than ruling rigid tied ones like companies. In companies, consent that legitimates powers and responsibility, economic resources and hierarchical relations all follow the same direction, from top to bottom, and each supports the others. In trade unions we have two opposite streams, from grassroots to top as far consent and, generally, most economic resources are concerned, and from top to bottom with regard to hierarchical rapports.

Becoming aware of this phenomenon often leads to frustration especially when striving to improve efficiency and set effective targets and rules for the entire organization. However such an attitude would be wrong. Efficiency can be improved by adapting to the structural characters of loosely coupled organizations rather than by struggling against them. First of all, there are rigid tied organizational areas within loosely coupled systems, which can be ruled and controlled more strictly than it can be done with the overall system. Local or federation structures and central national offices are, under several aspects, like companies and can be ruled in a similar way. As for the general organization, the professional competence related to organizational roles can be improved. And we must not forget that loosely coupled does not stand for absence of organization but defines a different form of organization. According to the different context there is a balance between the tension generated by the flow of resources and the flow of hierarchy, which generates the following four main organizational dilemmas.

The second structural character of trade unions as organization is that they have to face four organizational dilemmas. It means that there are requirements attracting the organizational solution towards two symmetrical opposite directions. The one best solution does not exist and every time the decision is posed in an intermediate position chosen according to the assessment of the existing conditions. This has two important consequences: the solution can never fully satisfy the organization needs and each solution will be temporary and local. In a loosely coupled system, thanks to the autonomy of components, different choices of balance between the two opposite sides of a dilemma are possible. This possibility confirms the great difference existing between loosely coupled systems and the theoretical image of a perfect mechanism which functions without any external influence. The four organizational dilemmas of trade unions are: influence vs. membership; sector aggregation vs. general aggregation; managers vs. feudal lords; officers vs. elected.

The first dilemma is between influence and membership. As above said, the ability of a trade union to attract and keep its members increases as its objectives and actions are successful in fulfilling its members’ perceived interests. However, workers’ differences in terms of sectors, skills, roles, age, family conditions, place of work, etc. acquire relevance when trade unions’ request go beyond asking for better salaries. On the other hand, trade unions need to concentrate on one or very few focal points in order to be incisive. Consequently, it is necessary to balance the need for one precise objective common to the entire organization and the opposite need for objectives which are perceived by members as corresponding to their specific interests. The different levels and local/sector structures of the organization are useful to solve this dilemma as they can emphasise different aspects of the same objective, communicate with the members to listen to their needs, problems, interests, understand their priorities and show the coherence between those and the organization's main objectives.

Such process confirms the usefulness of different levels, and various discussion and decision organs, which are first of all seats of internal communication where differences find a place to be expressed and eventually solved. The close relationship between a trade union and its members is a criterion to face the second dilemma: sector aggregation vs. general one, which is also called balance between the vertical and horizontal dimensions of trade unions. Vertical aggregation gives priority to the company sector where the workers are employed, whereas horizontal aggregation gives priority to the place where they work and often live. The solution is linked to the labour market conditions as well as the grade and dynamism of the economic development.

The two other dilemmas, concerning the full time trade unionists apparatus and the leading groups are very important. The dilemma about the role of full time officers is to what extent they behave like executives or managers of a company and to what degree they act like feudal lords. On the one hand, trade union officers perceive themselves - and/or are perceived – as guardians of the interests for the entire organization, a sort of spokesmen for the central leading group. On the other hand his feeling and his interlocutors' perception is as guardian of a specific structure's interests, spokesperson of a group of members and workers. The way a specific role is perceived does not necessarily depend on the behaviour and attitude of the trade union officer or leader covering a certain role. Interlocutors play an equally important role. In the same time and place, listening the same work by a trade union leader opposite perceptions of the role which that person is playing in that moment can occur. A group of workers may regard him as an executive of trade union organization while some officers or leaders of trade union upper organizational level as a feudal lord. Both groups expect the officer to show loyalty and feelings of belonging towards them. The ability to make interlocutors feel reassured in this regard is an important skill for full time officers and leaders and it can bring benefits to whole organization: not an easy balance to strike. 

Last, and obviously not least, the dilemma between officers and elected leaders. On the one hand they are mainly chosen for their competences in organizational activity, on the other hand they are chosen for their ability to obtain electors' consent. Trade unions need both profiles, which means that they may have to hire both competent full time officers holding high school diplomas or university degrees and elected leaders coming from trade union rank and file. However, this is not sufficient to solve the dilemma: Which collaborator profile has more impact and is more effective on trade union strategy? Every representative organization constantly needs to tackle these four dilemmas.

Italian confederations' structure and leading groups constitution

In Italy there are three main trade union confederations: CGIL (Condederazione Generale Italiana del Lavoro), CISL (Confederazione Italiana Sindacati Lavoratori) and UIL (Unione Italiana del Lavoro). Their story and differences are beyond the goals of this paper. I just remember that Italy has about sixty millions inhabitants and about seventeen millions of them are employed workers, plus almost six millions are self-employed ones. Table 3 shows the last data, regarding 2010. The figures indicate the declared members of each confederation, taken from their official websites. We can see that membership is still quite high, despite the decreasing tendency common to all the old industrialized countries. The trade union density, which is the percentage of unionized workers, is about 33%. It means that, on average, there is a union member every three white o blue collar workers: a relevant figure in a context of prevailingly small and medium enterprises. Besides, a very substantial part of the membership, about half of it, is made up of retired workers. They still remain members for several reasons, the main being the need of the trade union’s support and assistance in obtaining welfare state services and assessing fiscal duties. Their membership fees are very important to trade unions because the labour market and company organizations have deeply changed over the last decades. Trade unions' activities more difficult and expensive to carry out. There are no longer large factories and reaching, contacting, persuading workers - spread in several medium and small-sized enterprises - to join in unions cost much more time, energy, money. Furthermore uncertainty for job stability has increased employers' bargaining power. The resources coming from retired members are therefore quite important. At the same time a trade union confederation - strong for it representation among active workers - can more effectively rise its voice defending pensioners' interests.

Employed workers	17.077.000
Self-employed	5.734.000
Total employment	22.811.000
CGIL total membership	5.748.269
CGIL working membership	2.751.646
CGIL retired membership	2.996.623
CISL total membership	4.542.354
CISL working membership	2.340.490
CISL retired membership	2.201.864
UIL total membership	1.872.183
UIL working membership	1.296.318
UIL retired membership	575.865
CGIL+CISL+UIL total membership	12.162.806
CGIL+CISL+UIL working membership	6.388.454
CGIL+CISL+UIL retired membership	5.774.352

Table 4 shows the structure of the organization and each level of leadership formation, which is the same for all three main confederations. At the base we have the members who have the right to vote in two different structures, the plant joint union representatives and the local CISL federation leaders. Every three years members vote to elect the RSU, the workplace representative, together with non-member workers on lists presented by the different trade unions. Every four years the entire CISL holds a congress, a complex process that lasts a few months during which every level elects its leading bodies (general council and secretarial group, led by a general secretary) and its delegates to the upper level congress session.

Table 4. CISL organizational and operational structure

				Confederation Secretary
			Executive Confederation Committee
		National General Confederation Council
50% delegates	National Confederation Congress (every four years sessions)							50% delegates
National Sector Federation
		50% delegates
Regional Sector Federation							Regional Union Confederation
							50% delegates
Territorial Sector Federation							Territorial Union Confederation
		Workplace Representative
Members							Non-member workers

The process starts at grassroots level with meetings of members in every workplace, or grouped by zones when the companies are small and members are few. Members often know each other quite well, because Italian workers have the right, which was first won through bargaining and then, since 1970, generalized by law, which allows workers to meet inside the workplace during working hours for at least ten paid hours per year. After discussing the basic congress document and putting forward any proposals they would like to make, they elect the delegates to the territorial federation congress. There the debate is much more intense and usually lasts a couple of days, concluding with the election of territorial federation general council members, the secretarial group and the general secretary, delegates to the regional federation congress, which constitutes the following structure in the vertical organization’s line, and to the territorial confederation union congress, the so called horizontal structure, joint representation of all local sector federation. So the process follows parallel paths: one involves the sector federations, which constitute the so-called vertical line, while the other involves confederation and is called the horizontal line. Therefore, the delegates to the regional session of every territorial federation join in the regional federation session to carry out a new phase of the process. At the same time, all delegates elected by every territorial federation join in the confederation session, which is carried out in the same way: discussion and election of its leading organs (Secretary and council) and of delegates to the regional confederation congress. The process is repeated again for the national federation and regional confederation congresses. At this point there is a difference: half of the delegates come from regional federation congresses and the other half from territorial federation ones. The process concludes in the national confederation congress, which is made up of delegates elected both at national federation and regional confederations congresses, with the election of the national general confederation council and the national Secretary. At each congress the number of delegates is proportional to the number of members.

If it wants to be a truly democratic, representative organization, a trade union needs a true leadership selection process going from bottom to top. This is the best guarantee that each specific leadership, whether it be local, company-based or other, is as close as possible to its surrounding environment, in order to carry out its mission of representation. Of course it is necessary to reach a synthesis of such broad differences in order to arrive at well-defined, incisive decisions and proposals. This is the role of the leadership, too, but its relation with the base, down to the grassroots level, can never be as hierarchical as is that of management in business enterprises. Nobody is paid to be a union member, so there is no possibility for compulsory relations towards any member, freely associated. The union’s strength, and primarily its capacity to mobilize the base, comes from members’ effort, their commitment, and it depends on their will and free consensus.

The workplace representatives

Making comparisons in labour law and institutions is both difficult and dangerous. Difficult because labour institution and law are deeply embedded in national and local institutional, social and economic context. History, culture and values play a role in shaping each country law but the social context is much stronger in this case. Again they rise almost partly from the bottom, have to match the real conditions of work and life of a certain environment and time. It is dangerous because it is difficult to recognize what is similar and what is different, which is possible only with a good familiarity to the examined contexts. At the same time, comparison is useful because it helps us to become aware of gaps and weaknesses in our system and understand that certain elements we assumed as essential for the system to work are more the product of history than of necessity.

Having said that, we can go on to describe the Italian workplace representative.

In Italy the workplace employee representative is both a union's structure and a representative of all employed workers, even if non-members. It is called RSU, Rappresentanza Sindacale Unitaria; two thirds of the representatives are elected by all the workers in a company from lists of candidates nominated by the different unions, which have signed collective agreements enforced in that workplace – and one third of representatives are directly chosen by the same trade union organizations. The legislative basis for workplace representation is the 1970 Workers Statute, which provides for trade union representation at company level. However, although the law gives trade union representatives certain rights and protection, it does not provide detailed rules on how they should be chosen. The RSUs are regulated mainly by national agreements between workers and confederations of employers. An RSU can be set up when there are more than 15 employees in the workplace (five in agriculture); The number of RSU members varies according to the number of employees: 3 members for up to 200 employees, six from 201 to 500 employees and, thereafter, three extra seats for each 300 additional employees up to 3,000 employees; then three extra seats for each additional 500 employees. The key function of RSUs is to negotiate with the employers at workplace level. RSUs are intended to act as the workplace representatives of the trade unions and the agreements, which set them up, give them the power to negotiate binding agreements for their workplace as part of the bargaining structure.

The RSU does not have a major role in general trade union activity in the sense of promoting the union and union policy. The elected members who belong to a union can do it, but not the entire committee, which is usually made up of members belonging to different and sometimes competing trade unions. Larger companies may have organizational structures of trade union members who support a particular union, but this is more the role of the local trade union structure outside the workplace. Therefore, the company union does not keep any part of members’ fees, which go to the local sector federation. RSU members are legally entitled to paid time-off on the basis of a formula set out in the Workers Statute. This provides a total for the committee of an hour per employee per year where there are fewer than 200 employees; 8 hours a month for each 300 employees where there are fewer than 3,000; and 8 hours a month for each 500 employees or part thereof where there are more than 3,000. Each member is also entitled by law to 8 days unpaid leave a year for union business. In practice collective agreements often improve on this. Frequently the time-off will be re-divided so that leading members of the RSU have more time-off than others. In some large plants there may be someone completely freed from normal duties although this is less common than in the past.

The collective agreements are applied to all workers, members or non-members. So there is no direct benefit in becoming member. Some workers show a “free-rider” attitude and take advantage of trade union representation without bearing any individual cost. This attitude has been the centre of debate and the possibility of allowing only members to benefit from union services has been often taken into consideration in order to overcome the free-rider problem. I think that this would be wrong and dangerous. If we agree to differentiate between members and non-members by giving members a preferential treatment, I am afraid that, in many cases, we will run the risk that non- members, rather than members, will end up getting preferential treatment. As a matter of fact, employers do not welcome trade unions in the enterprise. Should a discriminating provision be legalized it is easy to predict that non unionized employees would benefit most from it.

In Italy union strategies to support unionization by means of differentiation practices relies on free services or special prices for members, such as legal assistance in work disputes, fiscal procedures, retirement plans as well as special discounts in purchasing goods and services. In the past, political ideology and sense of belonging played an important role in deciding to become a union member. Nowadays these values are being lost and there are more individual differences in the choice to join a union is motivated by different, more individual reasons, with variable degrees of belief, sense of belonging, use of services, friendship and trust in a particular local or company trade union leader.

The economic resources flow: the CISL case of members' fees.

The economic resources coming from members’ fees move in the bottom-top direction. Every organizational level with a permanent office needs economic resources to pay officers and operating costs. These organizational structures are legitimated by the congress, whose process we have seen above, and each has its own budget and free use of its own resources, within general CISL rules. This means that the principle of autonomy is also alive inside the CISL. Table 5 shows the distribution of funds to the various parts of the organization structure, each one with its own specific accounts. There are small differences among sector federations, which have the autonomy to decide the percentages that will be distributed to their different organizational structures. The example shown below is that of the metalworkers’ federation. If we take 100 to represent the members’ fees, 79.5% is distributed at the three federation levels: 65% to territorial, 6% to regional and 8.5% to national. The remaining 20.5% is distributed among the three confederation levels: 11.65% to territorial, 3.9% to regional and 4.95% to national.

By means of the trade union, workers find a way both as individuals and as a community to play a role on the political scene that does beyond the right of being parliamentary electors. In ancient times, politics was thought to express the pursuit of the common good. Then, in the Middle Ages, it changed its meaning and showed itself as the power to impose decisions, passing from the horizontal to the vertical dimension, becoming a hierarchical relation between superior and inferior. The democratic state offers a context within which these two dimensions can be reconciled, thanks to the pluralism that legitimates various social subjects in the decision-making process of the community. Civil society interacts with politics. As has been said above in interpreting the basic principles of the CISL, one guideline is that «the association or the social formation is the main way to achieve this difficult mediation, because it can transform individuals into persons and society into community, (…) if it is really based on individuals as persons, on their rights, their liberties, their powers, in other words if it accomplishes those democratic ethical values which cannot flourish in society at large if they are not born and do not flourish in the particular society»

Table 5. CISL distribution of members’ subscription fees (metalworkers fed. case)

8,5 %		4,95 %
National Sector Federation		National Confederation
6 %		3,9 %
Regional Sector Federation		Regional Union Confederation
65 %	20,5 %	11,65 %
Territorial Sector Federation		Territorial Union Confederation
100%
Members