IJPP ISSN: 2239-267X IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 76 Local development strategies for inner areas in Italy. A comparative analysis based on plan documents Gabriella Punziano P o st -d o cto r a l Re sear c h F ello w, Gr a n Sa sso Sc ie nc e I n st it ut e Vial e F. Cr i sp i, 7 , 6 7 1 0 0 - L’ Aq uil a, I ta l y gab r i ell a.p u nz ia no @ gs s i.it Giulia Urso P o st -d o cto r a l Re sear c h F ello w, Gr a n Sa sso Sc ie nc e I n st it ut e Vial e F. Cr i sp i, 7 , 6 7 1 0 0 - L’ Aq uil a, I ta l y gi ul ia. ur so @ gs si. it 1 KEYWORDS: Inner areas; Local development strategies; Policy design; Content and text analysis; National Strategy for Inner Areas (SNAI). 1 Sections “Introduction”, “Policy design and language: a constructivist perspective” and “Research Design” have been drawn up by Giulia Urso; sections “Methodology” and “Case studies: Preliminary local strategies within SNAI” have been finalized by Gabriella Punziano; Sections “Local development in plan documents: main findings” and “Conclusions” by both authors. mailto:gabriella.punziano@gssi.it mailto:giulia.urso@gssi.it Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 77 ABSTRACT Within the huge literature on local development policies produced across different disciplines, comparatively little attention has been paid to an important element as relevant as economic, financial and social capital: the cognitive element, needed in strategic thinking and complexity management, the “collective brain” guiding the decision-making process. In this paper, we investigate what we consider a direct “proxy” for this variable, which is supposed to incorporate that “usable knowledge” assisting those making policy choices: language. Language shapes the way problems are conceived, fixes priorities and delimits the range of strategic options. More specifically, our research question inquires which contextual factors are at stake in local development strategy design. The case studies were chosen among the pilot areas included in the Italian “National Strategy for Inner Areas”. Through a multidimensional content analysis of the plan documents available online, we explored the ways in which development strategies are locally interpreted. The techniques we used allowed us to make a comparative analysis, testing three effects that could have influenced local policy design: a geographical effect, a concept/policy transfer effect, and a framing effect. Broader, interesting reflections were drawn from research findings on the local embedded ability to designing consistent and effective development strategies. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 78 INTRODUCTION In the complex process of conceiving efficient local development strategies, besides the availability of economic and financial capital, a crucial role is undoubtedly played by the cognitive one (Dente, 2011). Especially in place-based approaches, the issue of knowledge is given much emphasis (Barca et al., 2012). Policy design acquires a paramount relevance as a premise for effectiveness. In principle, there needs to be a clear connection between the assumed causes of the problem being addressed, the actions put in place to solve it and an insight into what a desirable, feasible outcome would be. The fundamental argument behind this perspective is that “beginning with a coherent conception of the policy is likely to produce a more coherent result” (Peters, 2015: ix). Language thus plays a key role in policy design: it shapes the way we conceive a problem and thus fixes priorities, delimits the range of strategic options and thereby serves as a precursor to policy outcomes. In essence, “language has the capacity to make politics” (Hajer, Versteeg, 2005: 179). Hence, the importance our study assigns to language analysis and to the categories used to conceptualise action for local development. Fewer attempts have been made in analysing plan documents via their language in a comparative way. We are instead convinced that an interesting investigation in this field – which could help enrich the existing theoretical and empirical literature – must begin with a systematic examination of strategic documents, going through the texts describing objectives, actions, and expected results of the policy plan. Therefore, our research question inquires whether some (and eventually which) contextual factors are at stake in local development strategy design, observing a place-based, multi-level, centrally-driven policy targeting low-accessibility, peripheral areas in Italy, the “National Strategy for Inner Areas” (SNAI), within Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 79 which our case studies were selected. Their relevance for addressing the issue is twofold: i) rural, non-core areas are assumed to lack the necessary reserve of knowledge for effective policy- making (within SNAI); ii) a place-based approach (which informs SNAI), combining the efforts of both a better-prepared central level and less-skilled local governments, is supposed to mitigate this starting condition. The rest of the paper is organised as follows. Section 2 outlines the main theoretical groundings underpinning our research hypotheses, which are then described in Section 3. Section 4 and Section 5 present methodology and case studies respectively. Section 6 explains the main findings, and Section 7 concludes with some theoretical challenges for future research and a call for more practice-oriented analysis. POLICY DESIGN AND LANGUAGE: A CONSTRUCTIVIST PERSPECTIVE During the past four decades, public policies analysis has been a relatively fast-growing specialisation in the social sciences (Muller, 2009). It revolves around the understanding of the policymaking process, with the ultimate aim of bringing scientific knowledge to bear on decision-making (Fisher, 2003). Several approaches were developed to understand how policy is both made and implemented (Coletti, Urso, forthcoming); our focus is on the former aspect of the process, i.e. policy design. This strand of studies puts the emphasis primarily on the ideas informing a plan (Peters, 2015) and the framing of possible policy interventions (Stone, 1997; Chong, Druckmann, 2007), basically interpreting policies as social constructions. In this constructivist view, the more interesting one for the purposes of our research, policy design is shaped through argumentation (Fisher, Gottweiss, 2012). In the geography and Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 80 planning literature, the social-constructivist perspective was embedded into the shift to a relational approach that recognises place as a social construct, generated as meanings are given in particular contexts to particular spaces (Allen et al., 1998; Graham, Healey, 1999; Thrift, 1996). Most importantly, in developing understandings and dreams about the future of places, this approach recognizes that the imaginative content of strategic spatial planning episodes inevitably organizes the way the materialities are thought about (Healey, 2004: 49). As Healey (2004) underlines, articulating the vocabulary for a spatial strategy is essentially a political process, involving struggle and selectivity, not just between different interests and power blocs but also within the terrain of the mode of analysis and representation of the policy contents. Policy vocabulary reflects the institutional and cognitive context where it was produced. In local development strategies, in particular, it is a primary vehicle to understand the place qualities and the drivers for growth chosen by the area – that is, the uniqueness and specificity of the local project. In fact, many studies interpreted language as a constitutive force of policy, but they primarily focused on the role of rhetoric and narrative in the exercise of power by key actors, especially in Foucauldian-inspired research (Edelman, 1985; Fisher, 2003; Lees, 2004; Jacobs, 2006). Such an analytic attempt brings the question of the role of language used in policies – and especially of its performative power (Austin, 1962) – to the fore. Though it lies at the heart of policy-making and this is probably the field in which its performative character is most evident, language as a key construct in constituting, enabling or, in some cases, weakening coherent interventions has not been sufficiently theorised and empirically evidenced. What Healey (2004) noticed Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 81 about the use of the notions of space and place holds more generally for research on spatial planning. While much of the debate has concentrated on process and actors – and the related question of mobilisation and power (Fischler, 1995; Faludi, 1996; Neuman, 1996) – much less analysis has investigated the nature of some concepts being deployed. This is actually a rich research avenue, especially with reference to local development strategies, as the chosen trajectory of a place is continually contested through local struggles over meanings and values of place qualities. As a relational geography emphasises: “what are recognized as place qualities are shaped by multiple forces, producing multiple development pathways” (Healey, 2004: 50). This is why, in spite of the limitations of a rational, positivist view of the policy-making process (Fisher, 2003), a rigorous reasoning when combining problems, alternative solutions and their consequences in an organic text is undoubtedly needed. Language plays a vital role in building this strict internal logic and in imagining plausible scenarios. This also implies a deep contextual awareness of the issues tackled by policy, and a high level of intellectual capital to be embedded in decision-making. Information and knowledge are crucial in policy studies. Nevertheless, their optimal use is limited because of the following conditions, all of which affect the effective ability of policymaking to be rational (Lerner, Lasswell, 1951): bias in the conceptualization of the problem at stake, time constraint, uncertainty, incomplete information, complexity of chains of decisions in governance choices, and non-full predictability of consequences. Time constraint (in order to meet project deadlines) and a lack of local knowledge could have played a role in our case study. In this article, we examine the frames of reference and categories related to development used in recent local strategies produced in Italian inner areas, which explicitly mobilise endogenous Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 82 (cognitive) capital and especially local actors’ ability to make use of strategic thinking. The vocabulary of plan documents is under scrutiny, to find out how place qualities are conceived and possible futures dreamt about. Our research thus aids in developing an understanding of the multiple values and interpretations inner areas attach to the idea of their pathway to development. As discussed above, our study lies on the assumption that the way particular issues are discussed defines the way in which the topic is understood and the perceived possibilities to act. The concepts that are part of the local development discourse, which are part of the contextual “usable knowledge” (Lindblom, Cohen, 1979), are rooted in specific institutional contexts as well as cultural and political formations. By generating a specific vocabulary (Healey, 2004), they enable decision-makers to conceive a vision of the hindrances to development and legitimate certain choices but not others. The lack of connection between the main categories logically interrelated to one another could thereby have a disempowering effect in terms of strategic action in policy documents. Articulating local development problems – this not being a self- evident issue – requires conceptual frameworks and analytical capacities, as well as the creative force of agency sometimes lacking in local contexts. A critical aspect in local policies is that their knowledge base remains fragile and contentious and, arguably, locally-relevant information produced in the academic field and by the main organisations is not absorbed as a strategic resource in full. At the local level, knowledge is a scarce resource because of the limited scale, the small size of policymakers’ communities, and the supra-local nature of challenges to be addressed (Dotti, 2016). Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 83 RESEARCH DESIGN Having sketched the value of an analysis of language and the rationale for applying it to policy documents produced in peripheral regions, where the local stock of knowledge is often scant and dispersed, we now outline how this will be instrumental to our study and the contribution this could make to existing literature. As Peters (2015: 4) observed, the approach to design – we assumed in this paper, too – “emphasizes the need for comparative analysis”, which is particularly relevant also because policy design can be generated “by picking ideas from other settings, defined geographically or through policy domains” (Peters, 2015: 4). Our case studies are an extraordinary field of investigation in this sense: they are local strategies produced by different Italian areas that are part of the same multi-level strategy (SNAI) and are grouped together because of the similar start conditions hindering development (namely, low accessibility and distance from the main service provision centres). Pilot areas are asked to produce an adequate design within the same policy domain (local development), but are geographically differentiated, covering the whole national territory. Using all cases from SNAI allowed us to have comparable texts, produced by the same set of actors 2 , issued within the same policy-making process and drafted through a standard form. 2 SNAI has a complex governance structure, characterized by a horizontal and vertical coordination of numerous institutional public and private actors. Concerning the vertical dimension, all tiers of governments are involved: local Municipalities, Regions and central State. The local development interventions will be financed by all the available Community Funds (ERDF, ESF, EAFRD, EMFF), while the ones on essential services are to be financed through additional resources provided for under the Italian Stability Law. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 84 In order to unveil the “geography” of the vocabulary used in SNAI strategies, evaluate if some spatial patterns exist, and attempt to give some explanations for them, we presume the existence of three effects in policy design in inner areas: i) a geographical effect; ii) a concept/policy transfer effect; iii) a framing effect, which will be investigated through a multidimensional analysis of the contents of the local strategies produced within SNAI. Hypothesis 1 (H.1): geographically close strategic planning areas produce more similar local development strategies than distant ones. Research in the area of policy diffusion studies focused on identifying trends in timing, geography and resource similarities in the diffusion of policies (Evans, 2009). It was found that diffusion patterns emerge, among other dynamics, from geographical proximity of neighbouring areas (Berry, Berry, 1999). Diffusion also seeks to identify the patterns according to which policies spread and the geographic and structural characteristics of countries that might explain them (Stone, 2001). It evokes the idea of “contagion”. Methodologically, therefore, “geographic proximity is a frequent operationalization of both learning and emulation” (Gilardi, 2016). Nevertheless, it is also widely accepted that while offering a good starting point, the classic view of policy diffusion as geographic clustering is often limiting (Volden et al., 2008; Shipan, Volden, 2012; Desmarais et al., 2015). That being said, even when they are theoretically uninteresting or blunt, geography- based spatial lags are usually effective predictors of policies, as Gilardi (2016) puts it. He also adds that good measures for emulation are challenging and accurate direct indicators are still lacking, pointing at an indirect analysis through observable implications. We instead aim to seize it via a direct analysis of the content. In order to test H.1, aware that geographic contiguity represents nothing but a rough “neighbors variable” specification of Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 85 policy diffusion (Baybeck et al., 2011), we check if local strategies produced by pilot areas belonging to the same macro-area (North, Centre and South of Italy) cluster in terms of vocabulary used to develop them. Hypothesis 2 (H.2): local development strategies converge on the basis of the period of their online publication. As all documents are published online on the SNAI website 3 since their official approval, a “concept transfer” is likely to occur (Healey, 2004) with the borrowing of vocabulary from previous local strategies. New technologies make it easier to acquire information and disseminate them, thereby boosting learning from experiences (Nilsson, 2006). This is not negative per se, unless it takes the form of an indiscriminate transfer uncritically made by policymakers responding “to complexity by unreflectively cutting and pasting from foreign models” (Sharman, 2010: 623). This could weaken the performative action of a local development strategy, disempowering the foreseen policy interventions. Poor, partial transplantation is not as well-documented as the “success stories” in literature (Stone, 2012). Learning from early adopters is also widely recognised as a mechanism of policy diffusion (Shipan, Volden, 2008) and this seems to us a salient point dealing with strategies with different timing. Relying on the key point that when considering transfer one cannot study an iterative process without a temporal dimension (Marsh, Sharman, 2009), H.2 will be explored by adding the “time” element in our analysis, grouping local strategies by the period of their online publication. 3 http://www.agenziacoesione.gov.it/it/arint/index.html http://www.agenziacoesione.gov.it/it/arint/index.html Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 86 Hypothesis 3 (H.3): guidelines and formats provided by central government trap local strategic documents into some defined sets of categories and meanings. In line with what we discussed under H.2, SNAI acknowledges that inner areas are not fully equipped with all the economic and cognitive resources needed to carry out efficient development strategies (UVAL, 2014). Aiming at mobilising the “cognitive capital” of selected areas, central government assists lower levels in conceiving their pathway out of marginalisation, in a paternalistic attitude. Local strategies, although meant to be highly context- specific in SNAI’s intention, can be developed around a limited number of themed strands. In terms of policy design, coming back more explicitly to a constructivist view, it is worth highlighting that: rather than having a single answer to the policy problem different sets of argument ideas will be associated with different design, and may produce different types of intervention (Peters, 2015: 3). This framing issue is a crucial political process: it determines not only the kind of policy options to be contemplated but also the actors who will gain resources from the policy (Peters, 2015). This is why, as Healey (2004) asserts, the kind of imagination that is evoked by the vocabulary used in a strategic spatial planning “frame” does matter. Through our content and textual analysis of the development plan documents, we will assess whether the initial inputs (both in terms of eligible sets of actions and of drafting form) provided by the central level harnessed local strategies. In the next paragraphs, after briefly illustrating methodology and case studies, we present the main research findings under the three outlined hypotheses. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 87 METHODOLOGY Acknowledging the primordial relevance of the vocabulary used in policy design, in our study we examined it through a Content Analysis perspective (Benzécri, 1973; Stamler, 2001; Amaturo, Punziano, 2013; Tipaldo, 2014) aimed at directly exploring the contents of the local development plan documents produced by some of the pilot areas selected within SNAI. Content analysis can be considered as a wide and heterogeneous range of both qualitative and quantitative techniques for the interpretation of documents within communication or signification processes, be they written, oral, iconic, audio-visual, hypertextual or multimedia. We adopted a quantitative approach, treating text – in this view a set of “minimal units of meaning” – as a set of data comparable to numerical ones, on which statistical analysis can be applied after appropriate processing procedures (Bolasco, et al., 2004). A clarification is needed. This choice is not free from criticisms, limitations and pitfalls. As well-explained by Krippendorff (2004), a quantitative study of texts based on a multidimensional approach 4 – assuming that it is possible to investigate the emerging meanings from the texts and vocabularies documents employ without having an a priori knowledge of it – cannot but develop an exploratory and contextual knowledge, that is the “local optimum” (Fallery, Rodhain, 2007), which is valid one and only in that specific case. The absence of an a priori knowledge allows to study the 4 The multidimensional approach gives more emphasis to the structural rather than distributional assumptions in the analysis of relations among the investigated units – the single words in the vocabulary in our case (Han, Kamber, 2011). Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 88 relationship among texts and their structures and look at the most plausible explanations of the detected trends. A heavy but conscious bias is introduced, which influences the chance to find a particular set of answers, and no other, to research questions. A proper, deep analysis, taking these limits into account, should always be carried out using these techniques as an aid or as a tool (Krippendorff, 2004). However, when the main goal of a study is admittedly exploratory and aimed at comparing areas producing vocabularies related to a specific domain (local development), these approaches could be insightful being aware of bounds and biases. Through our expected results we do not try to obtain inferences or generalisations: we used them as support in the empirical demonstration of some linear theoretical hypotheses. In our case study, local strategies were investigated through a correspondence analysis on the text data. It allowed us to disclose: the emerging key concepts and the way they are used within the discourse; the strategies for which they are different or similar to one another; the focuses of local strategies, and, how the detected topics contribute to building spaces of meaning within the whole process of local policy-making. The ultimate goal is to assess how the use of a certain vocabulary in policy design gives rise to types and quantity of strategic options, helping us to test the three hypotheses at the heart of our research. The main expected results are the emergence of a first summary of the information contained in the strategies and the identification of the latent meaning dimensions underlying the different concepts of local development. We relied on three techniques, namely: Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 89 - Frequencies analysis 5 , to discover the peculiar language within the preliminary drafts; - Lexical Correspondences Analysis 6 , to compare the emerging latent meaning dimensions related to different spheres of discourse; - Cluster Analysis 7 , to classify the conceptualisations of local development expressed within strategies. CASE STUDIES: PRELIMINARY LOCAL STRATEGIES WITHIN SNAI For the purposes of our analysis, the texts prepared within the Italian “National Strategy for Inner Areas” (SNAI) are a stimulating case study as they are all public and available online. Launched in 2012, SNAI’s aim is to create developmental tools likely to foster a series of improvements in the wellbeing of populations living in inner areas (Urso, 2016), through both better access to basic services and use of the territorial capital 8 , defined within SNAI as the natural, cultural and cognitive capital embedded in a place. This should lead to an inversion of the 5 This procedure provides hints on absolute or relative frequencies of words, key- symbols or segments in a text or in a set of texts, as an indicator of interest and salience of meaning. 6 A factorial analysis technique on textual data devoted to produce a synthesis of the information contained in the analysed texts; a graphical representation of the network of associations between words and between words and texts; the connection between textual data and context data (Benzécri, 1973; Amaturo, 1989). 7 A dimension reduction analysis aiming at synthetizing the analysed information in few characteristic groups (Lebart, 1994). 8 Territorial capital is defined as the system of territorial assets of economic, cultural, social and environmental nature that ensures the development potential of places (Perucca, 2014). Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 90 negative demographic trend. By means of an initial screening of the national territory, 21 areas have been selected to enter the pilot phase. What is interesting from our perspective is the multi-step text production pilot areas are requested by the central government, which provides feedback and ensures support throughout, right up to its completion. It is a territorial co-planning operation carried out by the Minister for Economic Development and coordinated by a Technical Committee assisting local actors. Three kinds of documents are to be produced by local institutions before the complete package of interventions will be formally sanctioned through a Programme Framework Agreement: a brief draft of the strategy (Text 1); a preliminary extended version of the strategy (Text 2); the proper local strategy (Text 3). We examined twelve Texts 2 9 , which follow a precise scheme prepared by the central government, including the following sections: key idea; education; local development; health; mobility; municipalities involved; main actors; expected results 10 . In order to test our research hypotheses, we took into consideration the only paragraph devoted to local development. For this part, as for the other, pilot areas were provided with some guidelines for elaborating their strategy. This could lead to a potential bias because of the framing effects for which we checked. As we have seen, this is a very relevant issue because it could be difficult for local areas to detach from the proposed structure received by the central government in the policy design phase. We selected the 9 Authors of texts are not clearly indicated. 10 http://www.agenziacoesione.gov.it/it/arint/Documenti_di_lavoro/index.html Template for drafting preliminary strategies is the document called “Format di preliminare di Strategia”. http://www.agenziacoesione.gov.it/it/arint/Documenti_di_lavoro/index.html Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 91 documents officially approved and available on the website 11 to test also the existence of a concept/policy-transfer effect: under this hypothesis, the texts published first influence those that come later. To this end, we grouped preliminary strategies by their online publishing period: Table 1 – Selected pilot areas by on-line diffusion period Source: our elaboration. Case studies were also chosen to ensure the three Italian macro- areas were represented (four pilot areas for each of them) in order to test geographical effects in strategy production: Table 2 – Selected pilot areas by location Source: our elaboration. 11 Preliminary draft strategies (Texts 2) allow a representation of the whole national territory, differently from the proper strategies (Texts 3) that are not available in many cases, yet. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 92 Figure 1 – Case studies representativeness Source: Produced by G. Di Giovanni, Gran Sasso Science Institute. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 93 LOCAL DEVELOPMENT IN PLAN DOCUMENTS: MAIN FINDINGS The first step taken was to analyse the contents of the sections devoted to local development within twelve Texts 2 in order to identify common words and peculiar language. Figure 2 shows word frequency diagrams. The one on the left represents the general vocabulary emerging from all local development sections. The other ones display the vocabulary of each single text: common words to all strategies are marked in orange, the same colour as the general diagram; specific words detectable in single strategies are assigned different colours. The higher the frequency of the words or key concepts, the greater their salience and the interest on that specific issue. The most frequent word is thus territory, followed by local and area, configuring a first sphere of meaning related to the policy “scale”. A second sphere of meaning collects words concerning economic activities (production, services, chain, companies, network, management, resources) as well as main sectors (tourism, cultural, agricultural, heritage). A third sphere includes words pertaining to a broader interpretation of development, embedding a social dimension (young, offer, quality, enhancement, path, social). Analysing common language 12 , a greater contribution to general vocabulary is made by the strategies outlined by Basso Pesarese- Anconetano, Alta Irpinia, Basso Sangro-Trigno and Alta Marmilla. These are mostly Italian Central and Southern areas and are the later strategies to be published online. This leads us to reflect on the concept/policy transfer effect: we would have expected the 12 This was operationalised as the amount of common words of each strategy out of the total number of the more important words across all strategies. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 94 opposite result, with the first published strategies (by “early adopters”) laying the groundwork for the general vocabulary. Reflecting on peculiar language some foci could be identified: 1. Tourism, mostly sports, and promotion of natural resources, like woodlands and parks (Antola-Tigullio, Valchiavenna, Valtellina), and of the primary sector (agriculture and typical food products, fishing, etc.); 2. Tourism, but mainly cultural and heritage, and enhancement of local productions and artefacts, as well as of local services and sense of belonging (Casentino-Valtiberina, Basso Pesarese- Anconetano, Alta Irpinia, Sud Ovest Orvietano and Alta Carnia); 3. A more diversified economic and social development, aiming at young employment, competitiveness, attractiveness, innovation, especially with reference to the agricultural and local food productions sectors (Alta Marmilla, Madonie, Montagna Materana and Basso Sangro-Trigno). A tendency to cluster by macro area emerges. This could be explained by the geographical effect hypothesis. The separation between northern and southern areas is well evident, except for the second group, which collects strategies from the whole national territory. Acknowledging the reciprocity (cycling) relation existing between place and policy-making, we interpret this geographical impact in its narrow sense: similar local structural (physical and cultural) conditions in close territories could have generated more similar policy options. IJPP ISSN: 2239-267X IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 95 Figure 2 – Word frequency diagrams: peculiar language and common words Source: our elaboration. IJPP ISSN: 2239-267X IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 96 In order to investigate the framing effect hypothesis, we built a matrix with a set of binary variables assessing the presence of each of the five assets foreseen in SNAI guidelines with reference to local development (Table 3). The themed strands are territorial protection and management; valorisation of natural and cultural resources; renewable energy; agricultural and agro-food systems; know-how and crafts. We also added a sixth variable to record the presence of activities that are not indicated in the official frame. Table 3 – Pilot areas by presence of thematic assets Source: our elaboration. The strategies that developed more policy lines among the ones provided by SNAI belong to “southern Italy” (Montagna Materana, Madonie, Alta Irpinia). They also imagined some more extra options not included in the given list of assets (together to Alta Marmilla, and Basso Sangro-Trigno, the only in the “Centre”). They all fall into the third and fourth publishing phases. The remaining texts are more strictly bound to the institutional guidelines in designing their own local development strategy: a greater framing effect seems to be at play in this group. Because of the limits of this kind of quantitative analysis, however, other Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 97 factors not detectable through it could play a role. The choice of concentrating efforts on only a few objectives could be one, for instance. In order to better interpret these results, we tested the three hypotheses by adopting a multivariate and multidimensional technique: Lexical Correspondence Analysis was coupled with a Cluster Analysis, to investigate the “semantic spaces” contributing to the production of “latent dimensions of meaning” in each text. Figure 3 graphically displays the main results. The information contained in texts were summarised through a factorial analysis. The axes on the plan represent the intersection of two latent variables, which are two substantial dimensions contained in texts, but not immediately detectable through a mere reading. The horizontal axis (from left to right) opposes two different ways of conceiving the pathway to local development. On the one side, we have a “generic”, a-specific vision based on the tertiary sector (namely tourism, mobility, and services of general interest), with a low specificity of the mentioned local resources. On the other side, we can distinguish a context-specific vision, substantiated by a text that makes place qualities more explicit, also aiming at external recognition through brands and quality certifications coupled with an investment in advanced sectors (renewable energy). The vertical axis (from top to bottom) opposes two categories of resources: Cultural, Human, Social Capital and Natural, Environmental and Built Capital. From the comparison of the twelve texts through a Cluster Analysis, three different “models” of local development strategies emerged. IJPP ISSN: 2239-267X IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 98 Figure 3 – Lexical Correspondence Analysis on the twelve local development sections Source: our elaboration. This factorial graph summarising the results of the Lexical Correspondence Analysis (about 43,000 words, 7,000 graphical forms, the percentage of distinct word 15,7%, final vocabulary made up of 600 clean graphical forms with a cut-off frequency equal to 10, 65,6% of extracted inertia from the first two showed factors. A non-hierarchical Cluster Analysis has been conducted, cutting the dendrogram in three classes – representing respectively 29%, 28%, and 43% of total variance – with 56% of extracted inertia. IJPP ISSN: 2239-267X IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 99 Starting from the bottom-left quadrant, we find a group made up of all texts produced in the “first online dissemination phase” and in the “northern” macro-area, except for Casentino-Valtiberina (belonging to “Centre” and “second phase”). This group, which we named “Tourism-led development”, gather non-specific strategies, mainly focused on tourism, and natural and urban environment (with no reference to place peculiarities in terms of offer and attractions). This class forms a very cohesive cluster, collecting words connected to a mountain, sports, and accessible tourism, such as wood, forest, mountain, park, sport, disabled, care, gym, Alpine, promotion, mobility, tourism, sustainable, environment, landscape, hospitality, accommodation, traditions. The top-left quadrant collects policy documents belonging to “Centre” and to the “third publication phase”, at the intersection of an “a-specific vision” and a “cultural, human and social capital” axes. The model falling into it was defined as “Unspecialised heritage-led development”, mainly connected to historical resources. Many assets appear on the plan and it is not possible to detect the leading ones that local contexts focus on as drivers for their pathway to development. This class proposes, in fact, a rather multi/general-purpose strategy, or even an over-purpose one, with a generic reference to a broad conception of heritage. The aim of enhancing cultural resources emerges, but no specific activity to this aim is mentioned. Most relevant words are culture, history, art, creativity, museum, tourism facilities, foreigners, city, historical centres, little companies, itineraries, associations, municipalities, archaeology. The last group shows a remarkable difference from the previous ones. It includes local strategies produced by the “southern” macro- area within the “second and forth online issue periods”. It occupies the half plane of a “Context-specific vision” focused both on the Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 100 “Cultural, Human and Social Capital” and the “Natural, Environmental and Built Capital”. It could be defined as a “Place- led development”, mainly linked to cultural capital. The area proposes a knowledge/innovation-driven development strategy, giving great centrality to local actors and to their ability to control and internalise knowledge and external information. The outlined model relies on endogenous assets and their self-sustainability, combining the enhancement of well-defined context-specific cultural resources (related to food, events, museums, and handicraft) with the promotion of the (renewable) energy sector and small industrial firms. Most frequent words are renewable, energy, regeneration, biodiversity, leadership, local actors, food, local productions, events, eco-museum, music, school, community, young, innovation, public, private sector, research, international, productivity, learning, integration, urban space, information. This set of terms is very far from the language used in the other preliminary drafts, forming a totally independent space of meaning. Relying on these results, we can discuss our three hypotheses. As for the geographical effect, a clusterisation is evident on the plan (“North”, in the bottom-left quadrant, “Centre” in the top-left and “South” in the right ones), suggesting an influence of the neighbouring variable in the design of local development strategies. The interpretation of findings for the concept/policy transfer hypothesis is more controversial: while first, second and third issued strategic documents are similar to one another, fourth disseminated texts differ very much. We would have expected small incremental changes to be displayed, little detaching from the previous periods as it is for the three first groups circulating. The fourth issued group, entirely from Southern Italy, instead, shows some breakthrough innovations, using a totally different vocabulary, which does not seem to build on the previous ones. This could be indirectly related to an intrinsic cause, referring again Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 101 to a geographical effect: the difference in initial territorial conditions. Alternatively, it may be because of some intellectual autonomy facilitated by the availability of more time. Concerning the framing effect hypothesis, strategies on the right side show a greater capacity to enrich the guidelines and the template provided by national government: they are the only ones that developed the asset of “renewable energy” and which foresaw activities not included within the centrally-proposed themed strands. If coupled with the previous finding, one possible, tentative explanation could be that having embedded SNAI objectives and methods, they showed an ability to design local development strategies that “extend” the “physical” (format) and “cognitive” (list of contents to be included) scheme outlined by the central government. As we saw in the first paragraph (Sharman, 2010), the first experimental pilot areas, since they had no points of reference, produced strategies fitting more strictly within the provided frame. CONCLUSION Researchers increasingly recognise the important role of language in the policy arena. In essence, the main assumption is that language used by decision makers both shapes and limits what policies focus on and how they articulate strategic options. It is indeed a necessary constituent of the sequence of decisions and resource commitments characterising policy action. Any kind of problem is in the end “socially constructed” (Feindt, Oels, 2005), building on specific concepts and categories. Our study stems from a deep interest in the role of ideas, knowledge and competence in shaping policies and from the awareness of the fundamentally performative character of language in policy-making, which challenges us to rethink decision makers’ Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 102 responsibilities. We, therefore, compared local development strategies produced by a number of pilot areas included in the “National Strategy for Inner Areas” through a content and text analysis. More specifically, we put forward three hypotheses accounting for some potential effects impacting policy design in these peripheral areas: a geographical effect, a concept/policy- transfer effect, and a framing effect. We found that geographical and framing effects seem to have played some role in shaping local development plans, while the interpretation of the policy-transfer impact is less plain to see. Findings of our exploratory study merit some deeper analysis to answer thoroughly such complex research questions. However, these insights could help to enrich the existing empirical literature in comparative policy analysis, as well as the debate on policy design in the framework of multi-level governance. Many issues within the geographical and PPA literature, such as path- dependence processes, imitation resulting in mimetic isomorphism (Stone, 2012) of strategic documents and local conditions for effective planning, among others, still call for a more comprehensive understanding. More specifically, further research, relying also on qualitative methods, could investigate how consistently vocabulary of plan documents is used and fits with real contextual needs of strategic areas, or endogenous and exogenous factors affecting local policy design. From a practice-oriented perspective, an interesting avenue could assess which governance models and participatory mechanisms are likely to be more conducive to produce locally effective place-based policies. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 103 REFERENCES Allen, J., Massey, D. & Cochrane, A. (1998). Rethinking the region. London: Routledge. Amaturo, E. (1989). L’analisi delle corrispondenze lessicali. In E. Amaturo, Analyse des données e analisi dei dati nelle scienze sociali. Torino: Centro Scientifico Editore, 55-75. Amaturo, E., Punziano, G. (2013) Content analysis. Tra comunicazione e politica. Milano: Ledizioni. Bolasco, S., Canzonetti, A., Capo, F. M., Della Ratta-Rinaldi, F., & Singh, B.K. (2005). Understanding Text Mining: a Pragmatic Approach. In S. Sirmakessis (ed.), Knowledge Mining, Series: Studies in Fuzziness and Soft Computing. Heidelberg: Springer Verlag, 31-51. Benzécri, J.P. (ed.) (1973). L’Analyse Des Données (Vol. 1). Paris: Dunod. Austin, J. L. (1962). How to do things with words. Cambridge, MA: Harvard Business School. Bachtler, J. & Polverari, L. (2016). Foreword. In N.F. Dotti (ed.), Learning from Implementation and Evaluation of the EU Cohesion Policy: Lessons from a research-policy dialogue. Brussels: RSA Research Network on Cohesion Policy. Barca, F., McCann, P., Rodríguez-Pose, A. (2012). The Case for Regional Development Intervention: Place-Based Versus Place- Neutral Approaches. Journal of Regional Science, 52(1), 134-152. DOI: 10.1111/j.1467-9787.2011.00756.x. Baybeck, B., Berry, W.D., & SiegeL, D.A. (2011). A Strategic Theory of Policy Diffusion via Intergovernmental Competition. The Journal of Politics, 73(1), 232-47. DOI: Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 104 10.1017/s0022381610000988. Berry, F.S. & Berry, W.D. (1999). Innovation and diffusion models in policy research. In P.A. Sabatier (ed.), Theories of the policy process. Boulder, CO.: Westview Press, 169-200. Chong, D. & Druckman, J.N. (2007). Framing theory. Annual Review of Political Science, 10,pp. 103-126. DOI: 10.1146/annurev.polisci.10.072805.103054. Coletti, P. & Urso, G. (forthcoming, 2017). Methodological challenges for policy learning. In N.F. Dotti (ed.), Knowledge, Policymaking and Learning for European Cities and Regions. Cheltenham, UK; Northampton, MA, USA: Edward Elgar Publishing. Dente, B. (2011). Le decisioni di policy. Come si prendono, come si studiano. Bologna: Il Mulino. Desmarais, B.A., Harden, J.J., & Boehmke, F.J. (2015). Persistent Policy Pathways: Inferring Diffusion Networks in the American States. American Political Science Review, 109(2), 392-406. DOI: http://dx.doi.org/10.1017/S0003055415000040. Dotti, N. F. (2017). Abundant water, abundant knowledge: Cognitive patterns for policy change in Brussels’ water management system. European Urban and Regional Studies, November 16 (Online First), 1-18. DOI: 10.1177/0969776416677621. Edelman, M. (1985). Political Language and Political Reality. PS: Political Science & Politics, 18(1), 10-19. DOI: http://dx.doi.org/10.1017/S1049096500021247. Evans, M. (2009). Policy transfer in critical perspective. Policy Studies, 30(3), 243-268. DOI: 10.1080/01442870902863828. http://dx.doi.org/10.1017/S1049096500021247 Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 105 Fallery, B., & Rodhain, F. (2007). Quatre approches pour l’analyse de données textuelles: lexicale, linguistique, cognitive, thématique. In XVI ème Conférence de l'Association Internationale de Management Stratégique AIMS. AIMS. http://www.strategie- aims.com/events/conferences/7-xvieme-conference-de-l- aims/communications/2078-quatre-approches-pour-lanalyse-de- donnees-textuelles-lexicale-linguistique-cognitive- thematique/download Faludi, A. (1996). Framing with images. Environment and Planning B: Planning and Design, 23(1), 93-108. DOI: 10.1068/b230093. Feindt, P. H. & Oels, A. (2005). Does discourse matter? Discourse analysis in environmental policy making. Journal of Environmental Policy & Planning, 7(3), 161-173. DOI: http://dx.doi.org/10.1080/15239080500339638. Fisher, F. (2003). Reframing Public Policy: Discursive Politics and Deliberative Practices. New York: Oxford University Press. Fisher, F. & Gottweiss, H. (2012). The Argumentative Turn Revisited: Public Policy as Communicative Practice. Durham, NC: Duke University Press. Fischler, R. (1995). Strategy and history in professional practice: planning as world-making. In H. Liggett & D. Perry (eds.), Spatial practices: Critical Explorations in Social/Spatial Theory. Thousand Oaks, CA: Sage, 13-58. Gilardi, F. (2016). Four Ways We Can Improve Policy Diffusion Research. State Politics & Policy Quarterly, 16(1), 8-21. DOI: 10.1177/1532440015608761. Graham, S. & Healey, P. (1999). Relational concepts in time and Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 106 space: issues for planning theory and practice. European Planning Studies, 7(5), 623-46. DOI: http://dx.doi.org/10.1080/09654319908720542. Hajer, M. & Versteeg, W. (2005). A decade of discourse analysis of environmental politics: Achievements, challenges, perspectives. Journal of Environmental Policy & Planning, 7(3), 175-184. DOI: http://dx.doi.org/10.1080/15239080500339646. Han, J., Pei, J., & Kamber, M. (2011). Data mining: concepts and techniques. London: Elsevier. Healey, P. (2002). On creating the “city” as a collective resource. Urban Studies, 39(10), 1777-92. DOI: 10.1080/0042098022000002957. Healey, P. (2004). The Treatment of Space and Place in the New Strategic Spatial Planning in Europe. International Journal of Urban and Regional Research, 28(1), 45-67. DOI: 10.1111/j.0309- 1317.2004.00502.x. Ingram, H., Schneider, A.L., & DeLeon, P. (2007). Social construction and policy design. In P. Sabatier (ed.), Theories of the policy process (2nd ed.), 93-128. Boulder, CO: Westview Press. Jacobs, K. (2006). Discourse Analysis and its Utility for Urban Policy Research. Urban Policy and Research, 24(1), 39-52. DOI: http://dx.doi.org/10.1080/08111140600590817. Krippendorff, K. (2004). Content analysis: An introduction to its methodology. London: Sage. Lebart, L. (1994). Complementary use of correspondence analysis and cluster analysis. In M.J. Greenacre, J. Blasius (eds.), Correspondence Analysis in the Social Sciences, 162-178. London: Academic Press. Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 107 Lees, L. (2004). Urban geography: discourse analysis and urban research. Progress in Human Geography, 28(1), 101-107. DOI: 10.1191/0309132504ph473pr. Lerner, D., & Lasswell, H.D. (1951). The Policy Sciences. Palo Alto, CA: Stanford University Press. Lindblom, C. E. & Cohen, D. K. (1979). Usable Knowledge: Social Science and Social Problem Solving. New Haven: Yale University Press. Marsh, D. & Sharman, J.C. (2009). Policy diffusion and policy transfer. Policy Studies, 30(3), 269-288. DOI: http://dx.doi.org/10.1080/01442870902863851. Muller, P. (2009). Les politiques publiques. Paris: Presses Universitaires de France. Neuman, M. (1996). Images as institution builders: metropolitan planning in Madrid. European Planning Studies, 4(3), 293-312. DOI: http://dx.doi.org/10.1080/09654319608720347. Nilsson, M. (2006). The Role of Assessments and Institutions for Policy Learning: A Study on Swedish Climate and Nuclear Policy Formation. Policy Sciences, 38(4), 225-49. DOI: 10.1007/s11077- 006-9006-7. Perucca, G. (2014). The Role of Territorial Capital in Local Economic Growth: Evidence from Italy. European Planning Studies, 22(3), 537-562. DOI: 10.1080/09654313.2013.771626. Peters, B.G. (2015). Advanced Introduction to Public Policy. Cheltenham, UK; Northampton, MA, USA: Edward Elgar Publishing. http://dx.doi.org/10.1080/01442870902863851 http://dx.doi.org/10.1080/09654319608720347 Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 108 Sharman, J.C. (2010). Dysfunctional Policy Transfer in National Tax Blacklists. Governance: an international journal of policy, administration, and institutions, 23(4), 623-639. DOI: 10.1111/j.1468-0491.2010.01501.x. Shipan, C.R., & Volden, C. (2008). The mechanisms of policy diffusion. American Journal of Political Science, 52(4), 840-857. DOI: 10.1111/j.1540-5907.2008.00346.x. Shipan, C.R., & Volden, C. (2012). Policy Diffusion: Seven Lessons for Scholars and Practitioners. Public Administration Review, 72(6), 788-96. DOI: 10.1111/j.1540-6210.2012.02610.x. Stemler, S. (2001). An overview of content analysis. Practical Assessment, Research & Evaluation, 7(17), 137-146. Stone, D. (1997). Policy Paradox: The Art of Political Decision- making. New York, NY: W.W. Norton. Stone, D. (2001). Learning Lessons, Policy Transfer and the International Diffusion of Policy Ideas. CSGR Working Paper No. 69/01. http://www2.warwick.ac.uk/fac/soc/pais/research/researchcentres/c sgr/research/abstracts/abwp6901/ Stone, D. (2012). Transfer and translation of policy. Policy Studies, 33(6), 483-499. DOI: 10.1080/01442872.2012.695933. Thrift, N. (1996). Spatial formations. London: Sage. Tipaldo, G. (2014). L’analisi del contenuto e i mass media. Bologna: Il Mulino. Volden, C., Michael, M.T., & Carpenter, D.P. (2008). A Formal Model of Learning and Policy Diffusion. American Political Science Review, 102(3), 319-32. DOI: http://www2.warwick.ac.uk/fac/soc/pais/research/researchcentres/csgr/research/abstracts/abwp6901/ http://www2.warwick.ac.uk/fac/soc/pais/research/researchcentres/csgr/research/abstracts/abwp6901/ Punziano, Urso – Local development strategies for inner areas in Italy. A comparative analysis based on plan documents IJPP – Italian Journal of Planning Practice Vol. VI, issue 1 - 2016 109 http://dx.doi.org/10.1017/S0003055408080271. Urso, G. (2016). Polycentric development policies: a reflection on the Italian “National Strategy for Inner Areas”. Procedia - Social and Behavioral Sciences, 223(C), 456-461. DOI: http://dx.doi.org/10.1016/j.sbspro.2016.05.275. UVAL (2014). A strategy for Inner Areas in Italy: definition, objectives, tools and governance. Materiali Uval Series, 31 (Documenti). Rome: http://www.agenziacoesione.gov.it/opencms/export/sites/dps/it/doc umentazione/servizi/materiali_uval/Documenti/MUVAL_31_Aree _interne_ENG.pdf Short Authors Biography: Gabriella Punziano is Post-doctoral Research Fellow at Gran Sasso Science Institute, Social Sciences Unit. She is a sociologist with particular interest in methodological issues. Giulia Urso is Post-doctoral Research Fellow at Gran Sasso Science Institute, Social Sciences Unit. Her research interests include local development, rural-urban relationships, cultural events. http://dx.doi.org/10.1016/j.sbspro.2016.05.275 http://www.agenziacoesione.gov.it/opencms/export/sites/dps/it/documentazione/servizi/materiali_uval/Documenti/MUVAL_31_Aree_interne_ENG.pdf http://www.agenziacoesione.gov.it/opencms/export/sites/dps/it/documentazione/servizi/materiali_uval/Documenti/MUVAL_31_Aree_interne_ENG.pdf http://www.agenziacoesione.gov.it/opencms/export/sites/dps/it/documentazione/servizi/materiali_uval/Documenti/MUVAL_31_Aree_interne_ENG.pdf