Gender Budgeting in Italy: A Laboratory for Alternative Methodologies?

  • Francesca Bettio
  • Annalisa Rosselli


Gender Budgeting (GB) was introduced in Italy in 2001. Numerous GB initiatives have been carried out since at the local level, although in absence of any organised attempt to build a common methodology and coordinate the different experiences. Taking advantage of the diversity of the Italian experiences, in this chapter we assess the main methodological approaches that have been concretely tried out on the ground. We identify three main approaches, respectively account-based, policy-based and capability-based gender budgeting. Based on our experience, our exchanges with other experts involved in GB and the available records, we briefly illustrate the main advantages and limitations of each approach.


The story of Gender Budgeting (GB) in Italy seems to confirm the stereotype of a country where individualism and disregard for the rules prevail. Indeed, scores of GB initiatives have been carried out at the local level over the last 15 years, as we shall see later in detail, but none at the national level. In 2009 the parliament passed a law1 which strongly recommended GB to all public institutions and administrations. The recommendation, which had very few applications, was reiterated and reinforced more recently in legislation in ‘following which a special unit of the Parliamentary Budget Office and a group of experts were set up to coordinate and foster action.’There has been no organised effort to build a common methodology promoted by the Ministry of Equal Opportunity or other governing bodies. A number of reports on GB activities reveal some common features across the different approaches attempted, but the limited coordination of these is mainly due to the personal initiative of the gender experts involved, sharing their experiences at conferences and workshops.

As we argue in this chapter, this peculiar development of GB in Italy has had its pros and cons. A principal advantage has been the freedom of experimentation due to the absence of national guidelines and lack of pre-determined format. Experts and administrators had to start from scratch each time and exercise their ingenuity, since most international models—at least those easily accessible—had been conceived for entire countries and not for much smaller communities. A main disadvantage has been that lack of common methodology and organised reflection on past experiences has made progress slow. Moreover, none of the GB initiatives were mandatory, or integral to the routine process of approving the budget, nor indeed did they respond to strong demand from civil society. Therefore, most of them have been short-lived and were discontinued when the officers who had promoted them came to the end of their term and the new administration chose to ignore the work of their predecessors. This lack of continuity has meant that staff training on GB was difficult and projects had to rely heavily on external experts. After the fiscal crisis of 2010 when austerity set in and funds for experts became hard to find, GB came to a total halt practically everywhere just when it was most needed to assess the impact of budget cuts. This discontinuity, together with some methodological fuzziness, explains why there are very few instances in Italy where we can maintain that GB has become an effective and permanent instrument for change in the allocation of resources in favour of gender equality.

Notwithstanding these critical observations, GB has been useful in Italy. Among the observable benefits it has certainly increased the gender awareness of many administrators outside the small circle of equal opportunity officers. It has contributed to the development and progress of GB by clarifying basic concepts, introducing new tools and devising ingenious indicators for the context analysis of small municipalities or specific institutions like universities.

A Short History of GB in Italy: 2000–2016

GB made its first appearance in Italy in September 2000, when the authors of this chapter organised an international workshop in Rome, with the financial support of the Special Commission for Equal Opportunities, a department within the Prime Minister’s Office but now disbanded. Some of the international experts contributing to this volume (Diane Elson and Susan Himmelweit2) participated in the workshop and explained the main tenets of GB to a large audience of local administrators, trade unionists and academic researchers from all over Italy. Important members of the Cabinet also attended and declared their intention of introducing this new tool into the national budget. Unfortunately, a few months later, the general election brought a change of government. As Prime Minister, Silvio Berlusconi proved not to be particularly sensitive to gender equality and that conference remained the only occasion for a long time when the national government showed some real interest in GB. As noted, only recently has, a group of experts been appointed by the Italian Parliament to study the possibility of GB in Italy.

Following the workshop in 2000, a small group of local administrators, convinced of the importance and usefulness of GB, decided to start pilot projects in their own administrations. The first GB project involved Sestri Levante, a small municipality (Comune) in the province of Genoa in Northern Italy, in the winter of 2001.3 The Region Emilia-Romagna and the municipality of Modena followed, producing their GB reports in 2003. Since then, there have been a number of GB projects at the regional, provincial and municipal level. Badalassi (2014) reports that 24 Comuni, 27 provinces and 6 regions in Italy had carried out GB exercises by 2007, but the number is much larger by now. They are scattered throughout the country, with the exception of Sardinia. A regional Law4 accounts for the high concentration in Tuscany.

It is worth recalling that in Italy there are 20 regions, which are the largest local administrations. Altogether regions encompass 103 provinces, which in turn encompass 8000 Comuni, almost 70 per cent of which have fewer than 5000 inhabitants. These different levels of local government differ not only in size but also in the structure of their budgets, reflecting a complex division of tasks and responsibilities. About 60 per cent of a typical Italian region’s total expenditure is devoted to health services; about 30 per cent of a municipality’s total expenditure goes into social services and more or less the same amount goes into public transport and waste collection. Until 2014 Provinces were responsible for labour market policies and training programmes. In 2014, the role of Provinces was considerably reduced with most of the tasks and responsibilities previously allotted to them being transferred to the regions, the main exception being road maintenance and schools.

GB thus requires different approaches and areas of expertise according to the level at which it is carried out. For example at regional level, GB requires access to National Health Service data and a team of experts including some with medical/health expertise. At provincial level until 2014, GB would include assessment of training programmes and labour market policies and at municipal level, GB concerns social services, transport and environmental policies.

The spread of GB in Italy, in the first decade of this century, owes much to the favourable attitude of some feminist administrators and their awareness that something had to be done to re-address the many gender unbalances of their country. Italy is ranked among the countries in the European Union (EU) with the lowest gender equality, according to the European Gender Equality Index (EGEI). Its performance is above the EU average in one area only, namely health, thanks to Italian women’s long life-expectancy. In all other respects the situation is far from being satisfactory.

However, fundamental support to carry out GB projects also came from the EU. The European Social Fund (ESF), in the programming period 2000–2006, earmarked several million euros specifically for measures to improve gender equality and increase female participation in the labour market. Funds that could not be spent otherwise were available to local administrations able to show that they knew how to deploy them. GB, with its promises of improving the effectiveness, efficiency and transparency of public expenditure, gave politicians the opportunity to kill two birds with one stone. It allocated funds to increase gender equality, as required by the EU, and allowed politicians to justify and explain their choices to voters increasingly dissatisfied and disgusted by scandals and corruption. ESF resources enabled local administrations to:
  1. a.

    hire experts who combined knowledge of gender issues with training in economics. Even the largest administrations did not have the necessary competences in-house, since people in the equal opportunity division (usually women) often had no knowledge of economics and accounting, while people in the budget division (usually men) often knew nothing about gender issues;

  2. b.

    remunerate the effort of employees involved in the GB project who perceived it as an additional task which increased their workload despite not being mandatory;

  3. c.

    publish the results of GB exercises and showcase them publicly.

The support of the ESF lasted approximately until 2009. The subsequent ESF programming periods had different designs and rules, and the administrations that wished to continue GB practices had to rely on their own resources. This would not have been a serious problem if:
  1. 1.

    consensus had been reached on a common format for GB, clear, easy and inexpensive to replicate as well as appropriate to the responsibilities and resources (data, expertise) available to the administration carrying it out, be it a Region, a Comune or even network of small Comune to Comuni. For the aforementioned reasons this has still not been possible, with the consequence that interesting results of many pilot projects have not been used to build a shared methodology;

  2. 2.

    politicians were really interested in GB as an instrument for change; this would have required continuity, instead most of them saw it as a propaganda and self-promotion tool. Very often basic gender disaggregated statistics and comments were posted on the URL of the administration as evidence that gender issues were being tackled;

  3. 3.

    there had been strong pressure from civil society in favour of GB. Although there are signs that this pressure still is effectively being exercised in some places—above all in the universities—many activists and non-governmental organisations (NGOs) no longer include GB in their list of priority requests to local governments.


To wrap up this short reconstruction of Italy’s GB history let us note that no political organisation or women’s group has ever attempted a gender-sensitive analysis of the budget from outside the administration. What makes it practically impossible is lack of key information, aggravated by the fact that public budgets are not usually organised by programmes. The prevalent format is still line-item-based or structured around the units responsible for managing funds, which generally makes it difficult to identify precise allocation of funds. Hence results depend heavily on the level of aggregation and appropriateness of expenditure classification to the purposes of GB analysis. Suppose, for example, that it were important to measure funds devoted to home-to-school buses and the closest line entry were expenditure on city buses. In this case it would be cumbersome for someone outside the administration to separate out expenditure on school buses and the results could be inaccurate. This explains why, to use Elson’s classification (Elson 2003), the political location of Italian GB exercises has always been within government. With very few exceptions, these exercises represent gender auditing operations or ex-post analysis of the budget.5

A Common Protocol for Italian GB Exercises?

Our aim in this and the following sections is to highlight the main similarities and differences we detect in the GB exercises carried out within the country. The similarities can constitute the components upon which to build a common protocol for GB at each level of administration. We understand “protocol” as it is used in medicine, that is, a detailed written set of instructions to guide the GB analysis or to assist the practitioner in the performance of the procedure. As regards differences among GB exercises, in some cases they are more apparent than real, more a matter of how the results are presented than the way they have been reached. In other cases, they call for comparative assessment of the methodologies adopted, in terms of reproducibility, costs and actual usefulness for policymakers.

Our survey of GB experiences in Italy is rather selective, drawing primarily from reports that are (or were at some point in time) available online, the small number of academic publications which refer to them, and our direct involvement in some of these experiences. Selectivity is both a choice and the inevitable outcome of shortcomings in coordination and spreading of the findings.6

Starting with the similarities, let us recall that all the Italian GB exercises were carried out by the administration itself, often but not always with the help of external gender experts. Moreover, these exercises frequently followed similar sequences which can be broken down into three steps:
  • context analysis—identification and calculation of a set of demographic, social and economic indicators of gender equality and women’s well-being in the reference area;

  • analysis of local gender equality practices within and for the administration being audited;

  • gender impact analysis of the administration’s expenditure and taxation.

The salient differences among the various exercises are to be seen mainly in the third step, namely gender impact analysis, while differences in the first two steps are often not systematic.

Step 1: Context Analysis

Context analysis is the most developed part of any GB exercise, and by far its main component. It is also the part which has shown the greatest progress over the years, as can be seen by comparing some of the earliest GB exercises with the most recent ones. In some of the earliest projects or in those where the intervention of gender experts was insufficient, the availability of gender disaggregated data often drove the analysis, and not the other way round. The result was often a mix of conventional gender gap indicators (e.g. employment and unemployment gender gaps) and more unusual ones (e.g. gender gap between holders of driving licences7), with pages filled with graphs and tables that meant little to the external reader. Although convergence has yet to be reached on a well-defined set of indicators relevant for GB, some order has been introduced with the streamlining of the indicators for context analysis around a list of “capabilities” inspired by the works of Amartya Sen and Martha Nussbaum, following an approach which research on assessing development and inequality has largely adopted in the last decades.8 Examples of capabilities are: caring for oneself and for the others, living a long and healthy life, acquiring knowledge, among others.

It is worth noting that although the approach of identifying and grouping context indicators around a basic list of capabilities was also adopted in the development of the EGEI (EIGE 2015), the latter cannot be credited as influencing GB exercises in Italy, as it was not available when most GB exercises were carried out. In the future, however, the EGEI could be used to standardise large-scale GB exercises in the country, but only at the regional or national level, since the scope of municipal-level exercises is too limited and the data required for EGEI calculations are not available on such a small scale. A study which calculates EGEI for all the 20 Italian regions in the years 2005 and 2010 (Amici and Stefani 2013) has already shown promising results. For example, it made clear that no substantial progress in terms of gender equality occurred between 2005 and 2010. An interesting and unexpected result is that Lombardy, one of the most developed Italian regions, had one of the worst scores regarding gender equality in decision-making.

Progress has not, however, been shown in the choice of suitable benchmarks, which makes it difficult to interpret context indicators. The problem has often been solved by taking national averages as benchmarks, while politically agreed targets and standards have been used far less frequently.

Fragmentation has also implied insufficient political pressure to provide statistics capable of capturing crucial elements of gender equality and the care economy, like family friendly organisation of work or time allotted to specific tasks. A small Comune does not have sufficient clout to convince the national statistical institute to experiment with new kinds of data. Occasionally suitable proxies have been found for missing indicators. For example, it has become customary to use a specific age group breakdown of the population to construct a rough estimate of supply and demand of care services. Specifically, age groups under 17 and over 80 are taken to identify people who require care services, while those aged between 18 and 59 represent the “life-work reconciliation area” and those between 60 and 79 the “area in need for support”. The proportions between these groups can thus serve as a measure of the possible workload for the caregivers, while trends for the respective group can serve to project the workload into the future at the local level.

Step 2: Analysis of Local Equal Opportunity Practices

Analysis of local equal opportunity practices is perhaps one of the most satisfactory accomplishments of GB in Italy. We can think of no GB exercise in the country that did not include analysis of the equal opportunities practice of the administration under examination in terms of composition of the personnel, career opportunities, work-life reconciliation practices, procurement policies and gender empowerment. In this respect, many recent GB exercises have demonstrated a satisfactory degree of in-depth analysis (see e.g. Provincia di Siena 2013) so that developing a common protocol to be applied to all administrations would require no great effort. The resulting protocol might indeed be “exportable” to other countries.

The second step in the Italian GB exercises was also frequently devoted to identifying the expenses involved in directly and explicitly tackling gender disparities. These expenses include support for the gender “machinery” (e.g. funds available to the Equal Opportunity Committee) as well as specific actions (e.g. funds to promote women’s entrepreneurship). However, it is generally recognised that this type of expenditure usually accounts for a very small share of the whole budget, but the attention paid to it nevertheless signals awareness of gender issues on the part of the administrations, and their willingness to address them.

Step 3: Approaches to Gender Impact Analysis

The differences between the GB exercises lie mainly in the choice of approach to gender impact analysis. Differences start with the objectives ascribed to the GB exercises. Account-based and policy-based approaches target gender inequality and assess the impact of the budget in terms of the likely increase or decrease of gender gaps. The capability approach addresses the overall well-being of women and therefore targets all the expenditures having women as (main) beneficiaries. In practice, the result may often be the same, since addressing women’s disadvantaged conditions frequently implies reducing gender inequalities. In principle, however, this difference in targets exists and there are cases when it may give rise to ambiguities. Compare, for example, paternity leave provisions, which target men but favour a fairer distribution of unpaid work, with money transfers to caregivers whose beneficiaries are mainly women but which reinforce gender roles. If the objective is the reduction of gender gaps the choice is clear, whereas it may not be so clear if the objective is expenditure that benefits women.

Here we aim to single out clear-cut differences and similarities, although in reality some GB exercises borrow from different approaches, which may be a source of confusion. We identify three main approaches:
  • the account-based approach,

  • the policy-based approach

  • the capability approach.

All the approaches address the expenditure side of the budget while the revenue side has so far received much less attention. Arguably, however, the major differences among the three approaches are procedural and methodological. In the interest of clarity, we will analyse the three approaches in turn, stressing the differences rather than the similarities.

The Account-Based Approach

The account-based approach re-classifies expenditure with the ultimate aim of assessing the congruence between the professed gender-relevant objectives of the administration and its actual budget allocations. It builds on the proposal by Sharp (2000) to classify expenditure into three categories: (1) “Specifically targeted expenditures by government departments and authorities to women or men in the community intended to meet their particular needs”; (2) “Equal employment opportunity expenditure by government agencies on their employees”; (3) “General or mainstream budget expenditures by government agencies which make goods or services available to the whole community, but which are assessed for their gender impact” (Sharp 2000, Appendix 3).

The GB exercise originally carried out9 for the Marche region (Regione Marche 2006) can be seen as an application of the account-based approach. In the mid-2000s, when the exercise was carried out, Marche was a prosperous region with one of the highest female employment rates in Italy and one of the lowest gender employment gaps, thanks partly to a relatively strong incidence of female employment in manufacturing. In our experience, it is no coincidence that the regions that performed best in terms of female employment and other gender equality indicators were more likely to finance GB initiatives. The case of Marche is indicative in this respect. The tacit aim of such initiatives was to showcase outcomes for women as reflecting successful previous commitment on the part of the administration while reiterating commitment to further improvements. In short, the willingness of the administration to invest in such initiatives in part reflected actual commitment to gender equality, and in part amounted to a marketing exercise.

The Marche GB is notable for the attempt to experiment with innovative reclassifications of expenditure, each for a different purpose. It combines two main coding systems. The first system ranks expenditure items by assigning a higher score to those accruing to sectors and branches of the paid economy with higher incidence of female wage employment and/or female entrepreneurship. While such a criterion may be criticised on the grounds that it might end up encouraging segregation, the experts carrying out the exercise at that time felt that in the short run, increasing employment should take priority over de-segregation The second coding system classifies expenditure in terms of the gender distribution of the final beneficiaries. For example, under the second code (i.e. final beneficiaries) the expenditure on locally produced fire-fighting equipment was rated “neutral” since both men and women are likely to benefit from better equipped fire brigades. Under the first coding system the same expenditure received a mixed evaluation. On the one hand it was assigned a low score on the presumption that benefits for female employment would be limited in this male-dominated manufacturing branch. However, this was mitigated by the high score gained thanks to the expectation that female entrepreneurship might be favoured, since female entrepreneurs were surprisingly well represented in this branch. Another example is financial support for a private research institute. In this case expenditure received a low score on positive female employment spill-overs because of the low share of women in the staff while the score on the female share of the beneficiaries was high, since a core activity of the institute was training high-school teachers, among whom women are in the majority.

The classification combining the two coding systems serves two purposes: it increases gender awareness by providing a way of reading each expenditure in terms of comparative relevance for men and women. It also provides a preliminary assessment of the expected impact of budget allocations on selected gender equality outcomes (e.g. women’s equal labour market participation or gender fairness in the allocation of public benefits).The actual exercises carried out in Italy were geared primarily to increasing gender awareness, partly because the assessment potential of this methodology has limitations. Gauging the formal commitment of the administration to furthering the integration of women into economic and social life is a commendable ambition. However, finding assessment criteria that can be applied across all sectors, are cost-effective and feasible in a relatively short time, is not easy without resorting to ex-ante, highly subjective judgements. Barbara Bergmann expressed a similar concern when she reviewed the first exercises of GB more than ten years ago. In her words:

The “all-sectors” approach is a highly demanding one, and there have to be serious questions as to its feasibility in most contexts, even in the case of countries for which a considerable body of data have been collected and are available. Where time, energy, resources, patience and even good will are severely limited, it makes sense to concentrate on the most urgent issues, and what these are is usually no mystery. (Bergmann 2005, 137–138)

Going no further than reclassifying expenditure is a specific shortcoming of the Marche exercise. However, it also shares with other account-based GB exercises the difficulty in identifying priorities other than somewhat general ones. Suppose, for example, that regional governments devote extra funds to increasing women’s presence in the managerial ranks of private and public company boards in the region but make cuts to elderly care services for an equivalent amount. The simplest account-based approach based on the trichotomous classification proposed by Sharp would record a (perfectly compensating) transfer between expenditure favouring equal employment for men and women (group 2) to expenditure specifically targeted on women (group1). Nothing in the methodology of the account-based approach would help deciding if avoiding cuts to services should have priority over allocating more funds to enhance women’s empowerment. More sophisticated account-based approaches may make such problems less visible, but not remove them altogether.

The Policy-Based Approach

Implementation of the policy-based approach in Italy can be exemplified by the GB exercises carried out for the Modena municipality and for a group of small municipalities in the province of Siena (see Bettio et al. 2006).10

At the time of the GB initiative, Modena and Siena (both the municipalities and the larger Provinces) could boast a high level of women’s employment by Italian standards. In 2001, for example, the employment rate of women aged 20–49 reached 77 per cent and 70 per cent in the provinces of Modena and Siena, respectively, which was much higher than the average employment rate for this group for the whole of Italy.11 Moreover, welfare provisions were generous enough in Modena to warrant comparison with some Scandinavian countries. Again, therefore, willingness to finance GB initiatives came from local governments that had little to fear from comparative assessment of gender equality records.

In both cases the GB exercise explicitly referred to a set of policy objectives that are widely recognised as desirable from a gender perspective in academic and policy circles, ranging from closing gaps in employment, wages or political representation to easing reconciliation of family and work and reducing violence against women. Local-level indicators related to such objectives were then computed to set priority goals within a context analysis where local conditions were assessed against selected benchmarks. Next, key policies were identified for selected priority areas, selection being necessitated by limited resources for analysis. Finally, the core auditing exercise consisted of identifying budget expenditure items or programmes that could be expected to further progress in the key policy areas, based on the literature and on past records. For example, reconciliation was identified as a key policy area in Modena given that the local fertility rate was among the lowest in the country at the time of the exercise. Fertility was also a high priority area in the budgetary analysis of the Sienese municipalities. In the latter case, public childcare provisions were analysed in detail in order to gauge resources devoted to reconciliation, and suggestions were advanced to set “internal” targets so as to monitor progress, for example, not only rate of coverage but also opening hours per week. In this specific case the best values recorded by the participating municipalities served as benchmarks.

The advantage of this methodology lies in setting measurable objectives and clearly identifying the implementation policies. Its strong point is definition of the “roadmap” or, more simply, the process. A clear disadvantage is that identification of goals, policy and targets is neither simple nor easy to standardise. Political dialogue and participatory processes may help the administration to identify broadly defined goals, but which policies to follow is a matter for experts who are often needed to make discretionary or complex judgements. Implementation of GB can thus be costly, the quality of auditing may depend on the competence and judgement of experts and the related complexity may deter administrators.

The Holistic Approach: Enhancing Women’s Capabilities

The capability approach to GB builds on Amartya Sen’s approach to human development, which Sen does not identify solely with amelioration in the amount of resources available to individuals, but with the “expansion of the ‘capabilities’ of persons to lead the kind of lives they value—and have reasons to value” (Sen 1999, 18). Development is therefore a multidimensional concept where the cultural and emotional aspects of human life play an important role in addition to material living conditions. Moreover, since what makes life worth living for each of us depends not only on our individual tastes and needs but also on the relations we have with the other members of our society, this approach focuses on the historical and political context in which individuals are embedded rather than taken in isolation. The capability approach to GB thus claims to capture the multidimensional aspects of well-being and to emphasise the importance of relations with the physical and social environment. Far from dealing with an abstract “economic agent”, it aims to improve the quality of lives of “flesh-and-blood women and men” (Picchio 2015, 215).

This is indeed an ambitious goal, but we would argue that, ultimately, all GB exercises aim at enhancing women’s well-being since reducing gender inequality in its many aspects is often a pre-requisite for improving the well-being of women. Nevertheless, it is again necessary to stress the difference between a society where the living conditions of women are improved and a more equal society. One clear case of divergence between the objectives of gender equality and women’s well-being is when gender equality is measured in terms of gender gaps and men’s conditions are levelled down to women’s. This happened, for example, during the recent crisis in Europe and elsewhere, with the paradoxical result that levelling downwards reduced gender gaps and could be claimed to have enhanced equality despite the fact that both women’s and men’s well-being plunged. In this instance however, the seeming improvement in gender equality is an unintended consequence of having chosen the gender gaps to measure it—while alternative measures such as gaps with respect to a given target—say 60 per cent employment rate for women—would not suffer from this shortcoming. More importantly, perhaps, no policy deliberately pursues downward levelling.

Another case where pursuing gender equality may not coincide with pursuing improvement in women’s well-being is when men are at a disadvantage relatively to women and it is their situation which warrants improvement. In rich countries this is often observed for educational achievements, life-expectancy or health conditions, but these gaps in favour of women are rather infrequent.

In sum, while differences exist between the explicit goals of improving women’s capability and pursuing gender equality, they are infrequent or conditional on the way gender equality is conceived of and measured. We would therefore take issue with the way proponents of the capability approach to GB stress its virtues in opposition to other approaches: “traditional GA [Gender Audit] models are mainly focused on how public resources are used for goals of equal opportunities, gender equality or efficiency of public policies. Thus, firstly, they work in the space of income; secondly, they measure only public expenditure for gender equality or efficiency aims, rather than measuring the level of human development that this expenditure allows women to achieve” (Addabbo et al. 2010, 484). To the best of our knowledge, no GB exercise ever confined itself “to the space of income”; non-income dimensions like women’s time use and empowerment being one of the hallmarks of all GB exercises.

If then goals are not so dissimilar, does the real difference between the capability approach and more traditional GB approaches lie in the intended scope of the exercise? The capability approach is holistic and aims at encompassing all dimensions of well-being. Other approaches tend to start from the assumption that lack of data and resources as well as implementation costs inevitably force the scope of effective GB exercises down to selected dimensions of well-being. Again, however, the difference may be more seeming than real. As we shall illustrate below with reference to the Italian case, actual implementations of the capability approach also confined analysis of the budget to selected dimensions of well-being. Arguably, the real difference with other GB experiences was in the process that was chosen to select priorities, with the capability approach giving more weight to participatory processes or dialogue with the local administrators than to the judgement of experts.

In order to illustrate the advantages and limitations of the capability approach it is useful to identify the basic sequence involved in the well-being GB exercises. The first task is to identify a list of capabilities that, according to a fairly consensual view, can adequately reflect the overall well-being of a person, from basic capabilities like the ability to lead a healthy life, to more complex ones like the ability to participate in social life. Authoritative lists have been proposed (Roybens 2003; for a survey, see EIGE 2013, 145), but even those lists which might emerge unambiguously at some point may need fine-tuning to reflect the specific socio-economic context. To proceed with the argument, let us assume, nevertheless, that such a list can be identified. The challenge faced by a capability-inspired approach is how to link each capability with budgetary policy. Assuming, in the interest of simplicity, that the budget is organised by programmes, the challenge is how to link each relevant capability to the outlays foreseen for each programme, as well as to the way different departments in the administration under scrutiny participate in deciding and managing these programmes.

A specific tool has been devised to this end. Addabbo et al. (2010, 488) present a matrix which they applied to a GB study for the Modena province and the Lazio region and has been replicated in other cases. The columns of the matrix are the dimensions which make up the well-being of the population12 and the rows are the departments of the administration. Each element of this matrix is identified by two indices: the row-index i and the column-index j; and the impact of department i on capability j. Element ij of the matrix is zero when the budgetary decisions of department i do not have any impact on dimension j. For example, a programme against air pollution, managed by the public works department, has no impact on the capability of “being educated and trained”, but is relevant to the capability “living a healthy life”.

The second and third tasks involve, respectively, qualitative impact assessment—that is, gauging whether the programmes managed by a given department may be expected to have any impact on each of the chosen capabilities, that is, in what proportion the programme affects the different capabilities. Problems clearly arise when the expected impact is greater than zero for more than one capability. If each programme had an impact on a single capability, it would be easy to evaluate how much the administration spends for each capability by adding up the cost of all the programmes which affect that capability. But each programme usually has an impact on several capabilities. Likewise, each capability may be affected by several programmes. How then to evaluate how much is spent for what? There are two alternatives. Either no distinction is made and the same expenditure is counted several times, which is not very useful, or experts decide that the cost of a programme which is expected to influence capabilities A and B, is, say, 30 per cent spent for A and 70 per cent for B. Of course this would be a highly subjective estimate.

The total expenditure which has thus been assigned to improve a given capability is divided between men and women in the penultimate stage (fourth task). Since the aim of the exercise is to improve the well-being of women, this step is clearly necessary. Again, however, this division is rather arbitrary. For example, the cost of a programme to improve road safety was divided between men and women on the basis of the prevalence of women among accident victims (GenderCAPP, 2009). Thus if, say, women are 60 per cent of the victims 60 per cent of the cost of the programme would be assumed to benefit them, with the paradoxical result that if the share of female victims increased, the administration would appear to spend more for the well-being of women!

The final task involves calculating the share of total expenditure devoted to further each capability and using it to measure of the commitment of the administration to that specific dimension of well-being.

The methodological challenges of operationalizing the capabilities approach are all too apparent from this brief review of concrete attempts to implement the approach in Italy; and so are the weaknesses of some of the choices that have been made during these attempts. This notwithstanding, concrete implementation advanced analysis and methodology in certain respects. In particular, it must be credited for having brought under the spotlight the issue of coordination among different decision-making units of the local administration. The construction of the matrix of capabilities at the initial stage of capability-based GB exercises can, in fact, represent an effective tool to encourage dialogue and enhance interaction and coordination within local administrations. A non-trivial advantage is reducing the risk of local authorities offering provisions that have opposite effects on a given capability.


The brief review of GB experiences in Italy presented in this chapter was intended to be critical, and may therefore have under-emphasised the achievements. Italian GB initiatives have been undertaken primarily by local governments and on a discretionary basis. This local characteristic helped spread awareness of the gendered nature of budgetary policies across the country and across levels of local government. Also, the mushrooming of initiatives allowed for widespread experimentation. The discretionary nature of the initiatives meant that in some cases the institutions sponsoring them provided strong support, although not infrequently alleged support for furthering gender equality by means of GB practices may have masked opportunistic interests in showcasing on the part of local politicians.

The flip side of the coin was severe limitations in the scope, duration and effectiveness of the exercises due to small scale and insufficient continuity over time: most exercises depended not only on the political orientation of the incumbent local government but also on the personal favour of some of its members, both of which are inevitably short-lived.

However, our main proposition in this chapter is that “internal” limitations may have prevailed over “external” limitations. A crucial limitation is, in our view, methodological fuzziness coupled with and aggravated by lack of coordination among initiatives. In the absence of common standards and well-tested methodologies for GB at the local level, few exercises actually went beyond in-depth context analysis and gender auditing of personnel policy. Those which did look exclusively at the expenditure side were, at best, able to perform qualitative ex-ante gender impact assessment of selected policies and provisions. To our knowledge, moreover, there was no follow-up to verify whether and how GB initiatives actually managed to change the budgetary policy of the administrations that had promoted them.

At the same time, the GB initiatives were often costly given that the administrations promoting them had to hire experts to assist the local staff in conducting an unfamiliar exercise. With less than evident pay-offs (at least in the short run) and dwindling funds as the financial crisis progressed and EU changed its priorities, it is no surprise that the growth of local-level GB in Italy has lost much of its momentum.

If we are right in deducing that internal limitations account for mixed results more than external constraints and lack of resources, then sounder methodology, common standards and coordinated efforts are called forth to revive GB in the country. Our own contribution in this direction is the preliminary attempt we made in this chapter to assess the main methodological approaches that have been concretely tried out on the ground. We identified three main approaches, respectively account-based, policy-based and capability-based GB. Based on our experience, our exchanges with other experts involved in GB and the available records, we have briefly illustrated the main advantages and limitations of each approach.

Our bottom line is that effective GB needs to follow a common protocol and enhance its effectiveness through coordination across the different initiatives. The (common) protocol we favour includes context analysis aimed at carefully selecting just a few shared and feasible policy priorities. Moreover, it sees policy proposals and ex-post monitoring as a “must-do”. By this we mean quantitative monitoring of the cost-effectiveness of the provisions which are already in place and are expected to further the chosen objective priorities while also proposing new provisions to be subsequently monitored. On balance, what we called the policy-based approach comes closest to the protocol we favour. However, what we are advocating needs to go beyond experimentation in this approach to date and, in the process, borrow successful innovations from the other approaches.


  1. 1.

    Law 150/2009 recommends GB as the (sole) instrument to assess the performance of all public administrations in terms of equal opportunities. However, the law provides no further indications about methodologies, timing and responsibilities. In 2016 another law (Law 163/2016) added a paragraph which invited public administrations to model their GB on past experiences at the local level.

  2. 2.

    Other speakers included Rhonda Sharp and Haroon Akram Lhodi.

  3. 3.

    The reports of these very first experiences are internal documents not available to general public. The authors have a copy which can be made available upon request.

  4. 4.

    Regional Law 16/2009 of the Tuscany Region promotes GB in all the municipalities to assess the equal opportunity content of their policies. Piedmont has a similar law (Regional Law 8/2009).

  5. 5.

    One exception is the 2011 GB in the Region Piedmont which is an ex-ante analysis (, accessed on April 24, 2017).

  6. 6.

    As evidenced by the fact that even a recent report commissioned by the Italian parliament (Ufficio Parlamentare del Bilancio 2016), allegedly presenting the state of the art of GB in Italy, actually shows many lacunae and inaccuracies. For example, it neglects some important cases (Marche Region, province of Siena).

  7. 7.

    See Provincia di Trieste (2012, 77).

  8. 8.

    The Capabilities Approach methodology as developed within Wellbeing Gender Budgets is discussed elsewhere in this volume. For applications to the Italian case, see below.

  9. 9.

    The same exercise was also carried out elsewhere, in particular for the Pistoia province (See; accessed on April 2, 2017).

  10. 10.

    Modena is a medium-size city (around 180,000 inhabitants in the early 2000s) and provincial capital in the Emilia Region, serving a (once) rich manufacturing and food-processing district. Siena is a much smaller town (around 50,000 inhabitants) located in the Tuscany region.

  11. 11.

    In 2001 the average Italian employment rate ranged from 58.2 per cent for women aged 30–34 to 42 per cent for those aged 50–54.

  12. 12.

    The dimensions of well-being which make up the matrix come to nine: “being educated and trained”, “living in healthy, safe places”, “moving around in the territory”, “feeling safe”, “having access to resources”, “living a healthy life”, “caring”, “having leisure time, enjoying beauty” and “being informed”.


  1. Addabbo, Tindara, Diego Lanzi, and Antonella Picchio. 2010. Gender Budgets: A Capability Approach. Journal of Human Development and Capabilities 11: 479–501.CrossRefGoogle Scholar
  2. Amici, Monica, and Maria Lucia Stefani. 2013. A Gender Equality Index for the Italian Regions. Banca d’Italia Quaderni di economia e Finanza 190.Google Scholar
  3. Badalassi, Giovanna. 2014. Il Bilancio di genere. Accessed 19 Apr 2017.
  4. Bergman, Barbara. 2005. Gender in Public Expenditure Reviews. In Public Expenditure Analysis, ed. Anwar Shah. Washington, DC: World Bank.Google Scholar
  5. Bettio, Francesca, Simonetta Botarelli, and Annalisa Rosselli. 2006. Come si costruisce un bilancio di genere Linee Guida per amministratrici/amministratori comunali, Pubblicazione a cura del Servizio Cultura, Pubblica Istruzione, Politiche Sociali, Servizi alla persona dell’Amministrazione provinciale di Siena, Siena.Google Scholar
  6. EIGE (European Institute for Gender Equality). 2015. Gender Equality Index 2015—Measuring Gender Equality in the European Union 2005–2012.Google Scholar
  7. EIGE European Institute for Gender Equality. 2013. Gender Equality Index, Report.Google Scholar
  8. Elson, Diane. 2003. Gender Mainstreaming and Gender Budgeting. Paper presented at the Conference Gender Equality and Europe’s Future, European Commission.Google Scholar
  9. GenderCAPP (T. Addabbo and S. Saltini eds). 2009. Gender auditing del Bilancio del Comune di Modena secondo l’approccio dello sviluppo umano, Accessed 19 Apr 2017.
  10. Picchio, Antonella. 2015. A Macro-Micro Classical Political-Economy Approach to Well-Being Gender Budgets. Politica Economica/Journal of Economic Policy 31: 155–174.Google Scholar
  11. Provincia di Siena. 2013. Le politiche di pari opportunità e Il Bilancio di Genere della provincia di Siena. Google Scholar
  12. Provincia di Trieste. 2012. Bilancio di Genere. Donne e Uomini per far crescere il territorio. 2nd ed. Accessed 19 Apr 2017.
  13. Regione, Marche. 2006. Il Bilancio di Genere. Esperienze e Percorsi di analisi gender mainstreaming della Regione Marche.Google Scholar
  14. Roybens, I. 2003. Sen’s Capability Approach and Gender Equality: Selecting Relevant Capabilities. Feminist Economics 9: 61–92.CrossRefGoogle Scholar
  15. Sen, Amantya. 1999. Development as Freedom. 1st ed. New York: Oxford University Press.Google Scholar
  16. Sharp, Rhonda. 2000. Gender Responsive Budget: The Australian Experience, International Workshop on Gender Auditing of Government Budget, Rome, mimeo.Google Scholar
  17. Ufficio Parlamentare di bilancio. 2016. Il bilancio di genere: un inquadramento generale, Focus tematico n.7, 28 ottobre.Google Scholar

Copyright information

© The Author(s) 2018

Authors and Affiliations

  • Francesca Bettio
    • 1
  • Annalisa Rosselli
    • 2
  1. 1.Università di SienaSienaItaly
  2. 2.Università di Roma Tor VergataRomeItaly

Personalised recommendations