Steering the Digital Agenda at Arm’s Length. All Wobble, No Spin: The Contextual Lens

. The paper uses a longitudinal case study of Italy’s digital agency to investigate eGovernment and a subject that hovers at the far edge of the academic radar: agencification , or the setting up of semi-autonomous organisations that operate at arm’s length from the relative ministry. The aim is to make a threefold contribution of international scope and significance to the eGovernment debate by mapping Italy’s chosen path to public-sector innovation. Framing the country’s digital agenda within the larger picture of ongoing New Public Management-driven administrative reforms, the authors assess whether mandating an arm’s length body to steer the eGovernment strategies at public-sector macro level has been successful. The structural-instrumental, cultural and environmental lens used to analyse the key contextual factors shows how the continuity and discontinuity that has shadowed Italy’s ICT policies can be blamed on shifts in leadership and diverse ideas of modernization; on the digital agency’s multiple, even conflicting mandates; and on the misalignment of the ‘original agency model’ with the public machinery’s embedded culture.


Introduction
"Context is messy. Dealing with context in order to explain the outcome of a political or administrative process means taking into account the decisions and actions by individual politicians or bureaucrats, the media's attention (or not) of (alleged or real) administrative malfeasance, ad hoc informal linkages between domestic and transnational institutions, and so on". Taken from a recent study by Jon Pierre [1: 42-43], this quotation forms the point of departure for this research thread.
The specific focus of the qualitative paper converges on two topics of considerable interest: eGovernment and agencification. In this work, the term eGovernment is meant as an instrument of public action for its potential to address public problems and affect a wide range of organizations and society at large, in terms of impact on the relationship between citizens and institutions and the internal working of the public sector [2]. Agencification is interpreted as the disaggregation of government departments into single-purpose agencies [3,4] that operate at arm's length. This design choice means that the ministries are responsible exclusively for developing the policies and that the agencies, while accountable to the ministry of reference, must operate under their own steam according to precise performance standards.
The article critically analyzes the uptake of eGovernment in Italy [5], using a longitudinal case study to respond to two research questions (RQ): -What institutional responses has Italy's government come up with to ensure the governance of its public-sector ICT policies? -What explains the continuity and discontinuity of those responses? The first, descriptive RQ comes from the accepted knowledge that showing what happened on the ground is an ineluctable step of any rigorous research approach, not least because 'theory building and theory testing … are themselves in part dependent on the availability of good descriptions' [6: 207]. The second, explanatory RQ refers to the 'long wave of Government innovation programs' [7: 254] and the agencification processes that have significantly shaped Italy's ICT policies and their outcomes since the 1990s.
Drawing on extant academic literature, official documents and the personal knowledge base of the research team of organizational scholars, the paper reconstructs the journey of a governmental agency created in 1993 to bring Italy's central administrations into the digital era, which, despite its short lifespan, has had to change both its name and position in the government machinery a good four times.
The 'structural-instrumental, cultural and environmental' are the three perspectives [8,9] that enable the article to make an original, tri-directional contribution to the eGovernment research. Above all, it highlights how eGovernment and agencification are enmeshed in a Napoleonic administrative landscape, a context that has been relatively underexplored by the international literature [10,11]. Second, it slots Italy's digital agenda into the larger scenario of political instability and NPM-driven reform waves [7]. In particular, the leadership shifts that have seen each new government impose their own modernization ideas on the country, the multiple, at times conflicting mandates given to the digital agenda agency, and the poor alignment of the 'agency model' with the culture embedded in the public machinery are the prime factors responsible for the continuity and discontinuity that shadows Italy's ICT policies. The article closes with an assessment of the national eGovernment strategies.
The article is organized as follows. Section 2 explains why the contextual lens was adopted to analyze agencification. Section 3 is dedicated to the methodological approach, while Section 4 frames the contextual backdrop of Italy's digital agency, charting its timeline, environment and effective level of success. The discussion presented in Section 5 reflects on the study's findings, which indicate that the myth perspective has ultimately succeeded over the government's arm's length, technocratic stance in giving momentum to Italy's eGovernment endeavour. Section 6 presents our final remarks, underscoring how the paper's two original features contribute to the current eGovernment debate.
The intrinsic features of ICT-related policies (including the strategic role of the State, the ongoing internal debate for development resources, the inherent multilevelstructure of administrative systems, the involvement of the citizenry, and the role of the private sector) are of considerable interest to both eGovernment scientists and policy scholars. Adopting a contextual approach to the analysis of the 'course and outcome' [8] of the processes of agencification in digital government generally means acknowledging the constraints and influences exerted by the political mindset, the institutional background and the environmental forces, three aspects that need to be analyzed using, respectively, a structural-instrumental, cultural and environmental lens [9].
On the whole, reforms can be perceived from a structural-instrumental standpoint as conscious organizational design or reengineering, given that the structure is used by the decision makers as an instrument to achieve objectives. But this requires preconditions that give the leaders a solid grip on the reform processes and that enable them to 'score high on rational calculation or means-end thinking' (see Dahl and Lindblom [12] cited in [9]).
The emphasis on values and models borrowed from private firms has been the 'guiding light' of Italy's public sector organizational reforms for the past 25 years, as evidenced by the creation of arm's length bodies, or agencification [3: 1] [8]. In theory, agencification is supposed to improve organizational capability by giving managers more freedom to manage. However, in practice, it may actually reduce management capacity within government departments, which, according to Andrews and colleagues [13], implies that 'there could be a positive or a negative relationship between the relative agencification of a public organization and overall capability' [13: 6]. Moreover, Höpfl [14] claims that agencification attempts to sharpen the distinction between the ministers 'driving the reform agenda … and civil servants responsible for performance, implementation (…) in a quasi-contractual 'public service agreement' or 'framework document' specifying the respective roles of 'sponsoring' departments and the chief executives of 'delivery agencies', whose performance is audited and measured, and who are in this sense accountable' [14: 42-43].
The cultural perspective sheds light on how reforms and change in public organizations trigger an institutionalization process that gradually introduces the 'core informal norms and values' that set the organizations on a path of cultural change and distinction. The fact that different countries and government institutions have different historical-cultural backgrounds means that their reforms follow a 'path dependent' course that gives each national reform a distinct complexion. The proposed reform of a public organization must be put to the test of 'cultural compatibility' [8]: 'reform initiatives that are incompatible with established norms and values in organizations will be rejected, while parts that are compatible will be implemented; controversial parts will be adapted so as to be made acceptable' [8: 132]. Hence, the reforms are likely to be more successful when their underlying values are more optimally aligned with the values embedded in the existing administrative system.
Public organizations are said to dwell in a dual environment: the technical part, which mainly focuses on efficiency, production and exchange, and the institutional part [15], which is more about issues such as the appropriate organizational structure, internal culture, recruitment policy, etc. The institutional environment is a breeding ground for the reform myths that develop, spread to other organizations and give the public organizations their isomorphic personalities [8]. These myths 'window-dress' the organization's image and increase its legitimacy. 'From a myth perspective, reform initiatives that correspond with current doctrines about 'good' and 'modern' organizations will gain acceptance more readily than initiatives that diverge from what is thought to be modern. The greater the correspondence between, on the one hand, problem definitions and suggested solutions in reform programmes, and, on the other hand, the circumstances of organizations perceived as well-run models for other organizations, the easier it will be to gain legitimacy and endorsement' [8].
The above concepts will be used below to read and interpret the case study.

Research approach
When the research questions are 'How?' and 'Why?', when the researcher has little control over the events, and when the focus is on what is currently happening in reallife contexts [16], the most indicated methodological approach is the case study. That is the method adopted here to respond to the research questions raised, which centre on how and why the agency in question has taken the direction observed and the role played by the peculiarities of Italy's public machinery. The research questions presented above are addressed through a longitudinal case study of the government agency responsible for Italy's digital agenda from 1993 to 2014 (which, for reasons of simplicity and consistency, given the several name changes this public body has undergone, we shall also refer to as the 'digital agency'). The 1993-2014 timeframe captures the setting within which the Agenzia per l'Italia Digitale 1 (AgID), initially called the Autorità per l'informatica nella pubblica amministrazione 2 (AIPA), was created and evolved; 1993 was the year in which Italy embarked on the substantive administrative reforms that developed into a 'permanent cycle of reforms' [17: 787] and, in 2001, the AIPA was mandated to implement Italy's first ever national eGovernment plan. However, more recently, the tide of administrative reforms has been brought practically to a standstill, victim of the financial crisis that has rendered the economic climate uncertain and significantly eroded the public resources available to drive change. The current government issued the latest measure, which gave the AgID a new statute.
The relevant strengths of the various data sources (mainly official documents, regulatory measures and archival reports) were identified by triangulating the data collection and analysis results and were used to corroborate the study's findings and conclusions. The research team examined the main organizational and operational implications of each of the agency's diverse configurations and administrative reporting systems in order to reliably map the nature and limitations of each permutation, even though, for reasons of space, only a small part of that information can be presented here.

From AIPA to AgID
Italy's first digital agency, the AIPA, was officially established on 12 February 1993-during a spate of particularly intense administrative reforms-to govern Italy's public-sector digital policies. It has since undergone a series of permutations that need to be set against the relevant events, background and environment in order to understand the evolutionary dynamics behind Italy's current digital agency, the AgID. Mapping that journey, means starting at the very beginning, in 1992, when the Italian cabinet appointed a public body, the Department of Public Service (DPS), to govern the ICT-enablement of the public administration, but without vesting it with specific powers of intervention to develop a digital agenda. Its job, in fact, was purely to advise the government on the best way to coordinate the online PA, to research and design ICT development policies and to issue circulars and proposals on possible interventions.
Hence, in 1993, the government created the AIPA, the country's first national digital agency. Ten years later, in 2003, the AIPA was replaced, by the Centro Nazionale per l'Informatica nella Pubblica Amministrazione 3 (CNIPA). In 2009, CNIPA was succeeded by the Ente nazionale per la digitalizzazione della pubblica amministrazione 4 (DigitPA). The DigitPA was then subsumed, along with the functions of two other public bodies, into the AgID, the latest digital agency created in August 2012.
Rewinding to 1993: the AIPA 'operates under the aegis of the Italian cabinet, has technical and functional autonomy and shall form its own independent opinions'. In fact, Italy's first public-sector digital agency was tasked with both the oversight of the public ICT market (reporting on the technical-economic congruity of the biggest contracts signed by the State administrative bodies) and the promotion and realization of large-scale infrastructure projects (such as the RUPA5 electronic PA networkcompleted in 2000 in accordance with the Department of Public Service's legislative and regulatory framework) or the building of networks to connect the information systems of the individual administrations. Other functions included monitoring public ICT programmes and producing annual reports for both the Government and Parliament. In early 1994, Chief Information Systems Officers were appointed at all the central administrations.
In 2000, external forces in the form of the global digital revolution and European momentum pushed the Italian government to sharpen its focus on ICT policies and approve the national eGovernment Action Plan. The aim was to give cohesive direc-6 tion to the various attempts to improve the quality of public service. This was the first of several eGovernment plans to develop a network that interconnected the information systems of both the local and the central administrations; to implement online public services; to set up two web portal systems, one for citizens and one for business; to integrate the registry offices' databases; and to promote the use of electronic identity cards and digital signatures. The federalist whirlwind that started to sweep Italy in 2001 (which lasted until 2007) then led Parliament to enact a special constitutional law, which put the local administrations (regional governments and municipalities) bang in the middle of the political action, making them the country's principal agents of change.
In 2001, the ruling centre-right party established the Ministry of Innovation and Technologies and its subordinate Department of Digital and Technological Innova-tion6, severing the administrative innovation policies from those of digitization. The appointment of the former head of IBM Italy, Lucio Stanca, as Minister of Innovation and Technologies was the clear opening of a door to the private-sector IT companies that previous governments had kept firmly shut. eGovernment thus became the means to get the different government levels to forge relations of effective institutional cooperation.
In 2002, the government issued a set of objectives that basically converged with those of the eGovernment Action Plan 2000.
In 2003, the downgrading of the AIPA to the CNIPA, with no change to either its institutional location or its mission, more or less put control back in political hands.
In 2004 and 2005, the lack of financial resources (budget cuts of more than 6%) put the brakes on Italy's digital agenda and sent public-sector ICT spending back to the 1995 level. The government then introduced the regional competence centres to not only recognize and diffuse best practices, but also to get the local administrations to transfer and share their knowledge and skills.
In 2005, the Codice dell'amministrazione digitale (CAD), the Digital Administration Law, gave a legislative anchor and regulatory compass to the country's eGovernment policies and machinery. Basically, CAD aggregates the norms in a similar way to the Austrian law that allows the federal government to define standard products in the ICT field [18]. Also in 2005, Italy's Finance Act gave the CNIPA the task of preparing framework contracts for the procurement of standard ICT services applications (e.g. computer protocols or disaster recovery solutions) and made their use obligatory for the public administrations, except in demonstrable cases of alternative solutions that better meet their specific needs.
In 2006, Prodi's centre-left government remerged the Ministry of Innovation and Technologies into the Ministry of Public Service (a child of the previous centre-right government), bringing them under the leadership of one minister, an arrangement maintained by the successive centre-right executive, which launched yet another PA reform. In 2009, the CNIPA was replaced with the DigitPA, which was given design, technical and operating functions (including technical consulting for both the Prime Minister and the regions, local bodies and other public administrations) but had to contest with the persistent scenario of growing economic constraints. The change of the eGovernment guard from the CNIPA to the DigitPA further centralized the Italian cabinet's control over the country's ICT policies.
In 2012, the technocratic Monti government came to power and instituted the AgID on 7 August. The AgID replaced and absorbed all the various functions previously split across three different public bodies: the DigitPA; the Agency for the Diffusion of Innovation Technologies;7 and the Department of Digital and Technological Innovation. However, the AgID did not inherit the DigitPA's role as sole operator of the PA network, which was transferred to CONSIP, the company set up by the Ministry of the Economy to manage the platform for the online purchase of goods and services. On 6 July 2012, the Monti government made e-procurement obligatory for the entire public sector.
The current centre-left government is increasingly turning the AgID into a technical creature of the executive. Indeed, the digital orchestra is conducted by a special steering committee chaired by the cabinet ministers' delegate; is responsible for identifying priorities of intervention and for monitoring implementation; and has the final say on the PA information systems strategic development model. The steering committee is made up of delegates from the Prime Minister's cabinet office, the Ministry of Economic Development, the Ministry of Education, Universities and Research, the Ministry of Public Administration, the Ministry of Economy and Finance, two delegates from the Conferenza Unificata Stato-Regioni8, and members of the newly formed Tavolo permanente per l'innovazione e l'Agenda digitale italiana9, an advisory body of innovation experts.

Outcomes
The eGovernment indices computed by the three transnational bodies of the UN, the OECD and the European Commission can be used to roughly measure and compare the outcomes of the combined actions of Italy's government policies and the interventions made by the Italian digital agency in its various guises.
The UN E-Government Development Index (EGDI) [19], based on a biannual survey, is split into three sub-indices: the Online Service Index (OSI), the Telecommunication Infrastructure Index (TII), and the Human Capital Index (HCI). As explained by the survey authors, the EGDI is not designed to track the development of egovernment in absolute terms but to enable comparisons between different countries. Of the meta-regions, Europe, while far from homogeneous, posted the best results in 2014, with France, Holland and the United Kingdom heading up the world's top ten, compared with Italy at 23 rd place, up from 32 nd in 2012. It is important to note that of the three EGDI sub-indices, the HCI (which basically measures literacy and education) shows only a minimal gap between Italy and the best performers of comparable size (France and the UK), indicating that Italy's low ranking is mainly the fault of the 8 other two indices, i.e., the TII and the OSI. The TII shows that Italy has a lower percentage of internet users (58% versus 83% in France and 87% in the UK) and fewer wired broadband subscriptions per 100 inhabitants (22 versus 37 in France and 34 in the UK), while the OSI reveals an even bigger divergence, with Italy's online service uptake standing at 67% compared with 88% for France and 79% for the UK.
The OSI result is especially significant because it concerns eGovernment itself, while the TII and the HCI can be seen as indicators of the preconditions needed to use the services. In fact, the survey built the OSI around six thematic sub-themes: wholeof-government, multichannel service delivery, bridging the digital divide, increasing usage, open government, and e-participation, which are precisely the areas that Italy worked hard to implement in 2012-2014, and that enabled it to leap nine rungs higher. Indeed, the OSI data of the earlier 2012 survey [20] shows Italy with 50%, France 77%, and the UK 85%. Nevertheless, that big improvement demands closer scrutiny in order to understand, for example, the higher degree of internet use (the so-called degree of uptake) at which the citizens and firms of Italy interacted with the public authorities in 2014 compared with 2012.
On the other hand, the OECD data for 2011 [21] confirms Italy's unsatisfactory performance compared to France, the UK, and the OECD average generally, indicating that only those Italian firms with more than 250 workers used the internet to expedite PA business (e.g., obtaining information, downloading and sending forms, or completing administrative procedures and case handling) to a degree closer to France and the UK, and above the OECD average. That is well below the internet uptake target of around 75% for all of Italy's firms, far lower than the OECD average (88%) and nowhere near the 90% of France and the UK. As the OECD itself points out, the index reveals a huge divergence in uptake between Italy's large and small enterprises (p. 154). Moreover, it underscores that Italian citizen uptake also fell far short of the OECD average in 2011, at approximately 25% versus 65% ( fig. 9.6, p. 157), a bottom-feeder in second-to-last place, beating only Chile.
The European Commission's 2014 data for Italy are similar to those of the UN. Notably, the Digital Economy and Society Index (DESI) 2015 [22] is a composite index that, unlike the EGDI, surveys not only a nation's eGovernment status, but also the overall digitization of both society and the economy. That said, one of the five indicators is Digital Public Services (DPS), which ranks Italy 15 th in Europe (with France 8 th and the UK 11 th ). Even though this result (0.42) was a tad short of the European average (0.47), it is the best indicator of all five of the sub-indices that make up DESI (which places Italy 25 th ). However, it is fairly meaningful that both the European Commission DPS index and the UN OSI, each designed to cover different dimensions, give similar rankings to all three countries.
In short, the transnational indices show that while Italy has improved its worldwide ranking and shortened the distance that sets it apart from France and the UK in the past two years, it still has much ground to cover. It should be noted also that the UN EGDI comprises indicators that are both complementary and "enabling" to eGovernment, which, as mentioned above, are more indicative of the preconditions needed to develop a digital government.
We can identify three interrelated perspectives -structural-instrumental, cultural and environmental -that have affected the steering of Italy's digital agenda. The idea of setting up new arm's length bodies (or agencification) can be seen as a process of organizational differentiation meant as the variation in formal structure and orientation from one institution to another to deal with the variety of tasks in uncertain environments. In the case of AIPA, this organizational design choice was a deliberate change in response to an emergent issue, i.e., the steering of the ICT-enabled transformation of Italy's PA. The then ministers had assumed that 'going digital' meant not only the computerization of the public processes and services, but also the rewriting of the ground rules: 'rule-based information technologies alter the template on which service activities are conducted, and facilitate the application of ever more powerful computing tools to services that are often fundamentally information based' [23: 135].
In fact, in the early 1990s, a period in which only a few Italian PAs had embraced the digital culture, the AIPA was empowered by both the regulatory framework and the need to concentrate and develop the capabilities required to launch and govern the ICT strategies. Adopting an instrumental view confirms that a cross-cutting issue perceived as 'not working' by the political leaders of the time helped create 'a readiness to look for new solutions' [18: 10], including the birth of a single-purpose digital agency. This favourable climate and an autonomous statute enabled the legitimization of the AIPA and weakened the sovereignty of the ICT providers. The infrastructural projects and the diverse technical standards that enhanced the capacity of the administrations 'to share and integrate information across both traditional and new organizational boundaries' [24: 7] helped to spur the modernization of both the PA and the country itself.
Nevertheless, a predominantly technocratic approach, heedless of the domestic administrative and institutional context, is likely to fail [18,[25][26][27]. In fact, as soon as the realization dawned that the AIPA projects also promoted greater uniformity between the State's administrative apparatuses, it sparked a situation of opposition and conflict [8]. The strongly embedded culture of the government ministries often succeeded in defusing the external pressures applied by the agency and its successors to drive change, and, not surprisingly, although the less controversial of the planned changes (or those more compatible with the prevailing political-administrative culture and the established routines) were implemented, other of the digital agency's projects remained just that, and never even saw the light of day.
In more recent years, the myth perspective whereby a better-organized public sector corresponds to an ICT-driven public sector has gained legitimacy and, as a result, spurred eGovernment in Italy. For example, in June 2014 it became obligatory for all Italian PAs and their suppliers to use exclusively electronic systems for the billing and filing of accounting documents. Interestingly, this reform package meshed with a broader transnational project, the EU's payment-integration initiative for the simplification of bank transfers denominated in Euro (Single Euro Payments Area, or SEPA). Seen from an institutional-environmental perspective, a favourable culture and support from the environment can propel digital transformation.

The different political takes on modernization
The coming into power of a succession of governments, each with their own particular idea of how to modernize Italy's PA and bring the country into the 21 st century, significantly increased the cast of ICT policy actors. The centre-left governments approached the reorganization of both the PA and the processes as a joint affair, mapping a long and complex journey towards the integration of the various back-offices. On the other hand, the centre-right governments saw the front-office services as a more immediate way to pluck the fruits of visibility and consensus.
The result of these significantly divergent approaches can be seen in the repeated merging and demerging since 2003 of the Dept. of Public Service and the Dept. of Digital and Technological Innovation, before both were ultimately abolished in 2012 and their mandates transferred to the AgID. The volatile institutional context has shifted and blurred the objectives and the priorities of the eGovernment plans, the implementation of which, given their nature and Italy's infrastructural shortcomings, otherwise would have been spread over the medium to longer term.

Multiple roles and motives of misalignment
Despite being officially established as an authority, in reality the AIPA combined the features of a ministerial department with those of an agency. The multiple roles attributed to the AIPA shrouded it in ambiguity [28]. Indeed, it was at the same time the regulator tasked with redressing the imbalances of an oligopolistic market dominated by suppliers, the 'watchdog' of the PAs and the ICT providers, to reduce opportunistic behaviour (and possible collusion), and the front-line market operator sent into the field for particularly important interventions. As Halligan says in his Australian study [29: 448], each of these roles "captures an organizational imperative that is externally grounded and usually has a basis in the agency's empowering legislation. Each has a different external driver, respectively: customers, clients, competitors, and politicians". The friction caused by cramming four models into one organization invariably sparks tensions and conflicts.
There are three interrelated reasons for this misalignment between the various digital agencies and the individual PAs. First, the objectives set out in the government guidelines are either incredibly vague or, to the contrary, fix on a specific detail. The difficulty of identifying the objectives has been compounded by the difficulty of assessing the ex-ante, in itinere and ex-post effects of the ICT-enabling projects. In fact, the 1990s saw the AIPA struggle to monitor both the spending and the number of ICT platforms and services purchased by the various administrations. Second, the negative economic cycle has ruled out the use of financial incentives to get the PAs to invest in technological and organizational innovation. Moreover, not all the administrations had the capabilities needed to launch large-scale projects or were even interested enough to measure the results of the initiatives implemented. Hence, many of the "zero-cost" innovations promoted by the AIPA et al. remained on the drawing board [30]. Finally, even though each ministry had been asked to appoint a Chief Information Officer (CIO) to specifically facilitate matters with each PA, the role was never fully recognized, which left the digital agency without the liaison officers, i.e., the unique points of reference, it needed to fulfil its mandate.
Interestingly, the EU has had a clearly observable impact. The momentum that spurred Italy's digital agenda in 2000 coincided with the launch of the eGovernment plans formulated to meet the targets set by the European Commission's Lisbon strategy. Seen from a myth perspective, these projects were part of a more general trend that the 'old-style' [25: xiv] governments of the time were eager to embrace as a way to gain international legitimacy.

Final remarks
The two original features of this paper are, first, the decision to investigate the combined effects of eGovernment and agencification at the public-sector macro level and, second, the adoption of a contextual lens [27] to study the evolution of an Italian public agency. Although the analysis of the arm's length body AIPA and its successors CNIPA, DigitPA and AgID has greatly enriched our knowledge of the digital government approach taken by Italy's political leaders, we will limit ourselves here to summing up the key findings. First, but by no means unexpected, is the fact that the single-purpose agency investigated in no way mirrors the ideal type advocated in the mainstream literature on agencification. Second, the digital agency had much more power to decide its objectives in the period 1993-1999, while the relative ministry's assessment of its performance and senior public managers was very low-key. After 1999, the AIPA and its successors gradually lost their autonomy and relevance, taking on more the guise of a government operational arm. Third, the article reveals a considerable rhetorical dimension among the many variables that shape the course and outcome of Italy's eGovernment policies [31], as well as the inability of the independent agency mandated to steer and reinforce the digital transformation of the public sector and, thus, to catalyze major change. Indeed, the actual steering action is so diluted that it bears no resemblance at all to the ICT-driven change boldly promised by each new Italian government, no matter their political affiliation.
Looping back to the first RQ 'What institutional responses has Italy's government come up with to ensure the governance of its public-sector ICT policies?' we can say that the proliferation of the agencies was originally driven by the absolute strategic approach taken to the institutional design implementations from 1999 to the present. The analysis has shown that Italy has done very little to improve the effectiveness of its eGovernment policies. At the same time, the regulatory framework used to govern the public-sector ICT strategies has not only grown unwieldy, but places too much emphasis on the technological, instrumental aspects of digital government, ignoring the bigger picture of digital citizenship, to which all the governments in question have paid only fleeting attention.
To respond to the second RQ 'What explains the continuity and discontinuity of those responses?' the main variables were analyzed using a theoretical lens sensitive to contextual factors [1,32,33]. Applying this interpretive key to the Italian case shows that the wobbly path of continuity and discontinuity that has taken the spin out of Italy's digital agenda can be attributed to specificities of a structural, cultural and environmental nature. What unique features were decisive in shaping the events that in many ways recur around the world? No general answer may be attempted here, even though, in a country like Italy with a decidedly Napoleonic administrative landscape, the absence of a permanent strategic body at the highest level is certainly a primary factor. The weakness shown in the early formative years has stunted the evolution and scope of action of the digital agency, which found itself in the unenviable position of having to run a strategically decisive process in the life of the country with a toolbox consisting of not much more than the power drill and screws needed to put up a framework of technical rules and regulations. However, the pre-eminently cognitive aim of the paper means that the analysis makes no attempt to correlate or map the dynamic interaction of the identified explicative variables, although we plan to redress that limitation in the next phase of this fledgling research path.