Obstacles to local cooperation in fragmented, left-behind economies: an integrated framework

Fostering cooperation among local stakeholders is a core aim of place-based policies, and it can generate economic benefits and help restore a sense of agency in left-behind communities. However, relatively little is known about how to induce local cooperation in low-trust, institutionally weak areas. This article develops an integrated theoretical framework to help diagnose the precise obstacles to cooperation faced in different types of adverse settings. Such a diagnosis can help design tailored local-and macro-level policies to address the obstacles to local cooperation. The utility of the proposed framework is demonstrated using a medium-n comparative case study design.


Introduction
Cooperation among local firms, as well as between firms and other local stakeholders, is often a key component of strategies through which left-behind places can carve out new economic opportunities in the globalised knowledge economy.In areas with fragmented business ownership structures, cooperation is a way for small-and mediumsized enterprises (SMEs) to collectively acquire capabilities that are unattainable for them individually, mitigating the productivity disadvantages that stem from their size (Asheim et al., 2011;Becattini et al., 2009;Burroni and Trigilia, 2001).Cooperation among private, public and civil society stakeholders is also a core feature of place-based policies, which emphasise the importance of local knowledge in providing tailored solutions to the problems of left-behind places (Beer et al., 2020).Beyond its economic benefits, local cooperation can help restore a local sense of agency and community, addressing the 'disruption of place attachments' that tend to accompany local economic decline (MacKinnon et al., 2022: 46).Despite these potential benefits, it is known that cooperation is often difficult to achieve, particularly in settings with low trust and weak institutions (Ferguson, 2013;North, 1990;Putnam, 1993).This difficulty makes it par-ticularly challenging to design and implement successful place-based policies in left-behind areas, which frequently suffer from 'a lack of civic assets and community facilities' (MacKinnon et al., 2022: 42).But precisely what is the nature of the obstacles that sometimes make even the most economically sensible forms of cooperation appear beyond reach?Addressing this question is necessary both for designing local-level strategies to trigger cooperation and for constructing macro-level policies that can facilitate local cooperative efforts in adverse circumstances.
This article analyses the obstacles to local cooperation in a setting that is adverse to the emergence of cooperative activities, namely, Greece.Greece is considered to have 'low levels of social capital and capacities for collective action', reflected in low shares of the population agreeing that most people can be trusted in large-scale surveys on interpersonal trust (Paraskevopoulos, 2007: 15).Nevertheless, local cooperation is important for the country, which heavily relies on SMEs, and which had to switch from its traditional, statist development model to export-oriented growth in the context of an economic crisis of unprecedented magnitude.In some exceptional cases, private, public and civil society stakeholders in specific areas came together and were able to collectively upgrade quality, establish a brand name based on place and innovate, thus navigating the transition to exportoriented growth successfully.This did not happen in other, similar areas, where cooperation could have had comparable benefits but proved too challenging to achieve.The paper leverages this variation through four pairs of matching local case study comparisons, which provide a unique glimpse into the obstacles that must be overcome for cooperation to emerge in adverse settings.
The article puts forward an integrated theoretical framework for analysing the obstacles to local cooperation in low-trust, left-behind places.The framework accounts for two types of obstacles that must be overcome sequentially for local cooperation to emerge in unfavourable settings.As emphasised in the sociology literature, a set of cognitive obstacles to cooperation must be addressed first.Particularly in cases where cooperative activities entail a high degree of innovation, cognitive obstacles such as entrepreneurial discovery and social fragmentation can prevent boundedly rational local actors from even conceptualising the full range of available cooperative strategies.Once these cognitive obstacles are overcome and the contours of potential cooperative activities become clear to most stakeholders, a set of collective action problems arise, as suggested in the classic political economy literature.Each type of obstacle requires distinct sets of solutions at the local and macro levels.
By bringing together the insights of two theoretical approaches in an integrated, sequential framework about the obstacles to local cooperation, this article contributes to our understanding of the nature of the left-behind problem, particularly in areas with fragmented ownership structures.In such areas, cooperation is an important ingredient of initiatives to help local firms switch to novel models of production, which can improve local prosperity in the context of the globalised knowledge economy.The framework presented in the article can be used as a diagnostic tool to help design place-based policies that address the obstacles to local cooperation.Empirically, the article highlights the potential utility of drawing on the experiences of former and current left-behind areas in countries outside the global core, some of which have faced similar challenges as left-behind places in highly advanced economies in the past.
The article is structured as follows.The next section situates this research in the literature on place-based policies for left-behind areas and develops the article's theoretical framework.The third section presents the article's empirical setting and methodology.The following two sections apply the proposed theoretical framework to analyse the specific obstacles to cooperation that were faced in areas with four distinct economic specialisations.Section four focuses on the collective action problems that hindered cooperation in areas specialising in the production of established goods and services, while section five highlights the importance of cognitive obstacles to cooperation in cases where higher degrees of innovation were required.The sixth section concludes.
Collective action problems, cognitive obstacles to cooperation and their relevance for left-behind places: the theoretical argument If place-based approaches aim at 'promoting economic productivity, generating creative policy responses and harnessing civic engagement' (Beer et al., 2020: 18), cooperation can play a central role in advancing all three goals.Starting with the former, despite their differences, the literatures on industrial districts, industrial clusters, regional innovation systems and innovative milieux share a focus on 'cooperation, networks, institutions, trust, interorganizational learning and knowledge transfer' as key elements of successful local development models in areas that heavily rely on SMEs (Asheim et al., 2011: 878).By working together, SMEs can collectively acquire capabilities that each firm could not afford individually, including for quality upgrading, branding, conducting R&D, innovating, satisfying bigger orders, training workers, accessing customised services and providing other club goods that are useful to the production process (Bianchi, 2001;Dei Ottati, 2002;Díez-Vial and Fernández-Olmos, 2013;Hancké, 2011;Muscio, 2006;Ornston, 2012).These capabilities can contribute to addressing the productivity gap between small and large firms.Local stakeholders can also come together to attract resources to their community and enhance the quality of the local residential and business environment, improving firms' ability to attract employees (Avlijaš, 2022).These factors are likely to influence the 'place-based "competitiveness" residual effect' that explains subnational geographical variation in economic performance once industry specialisation has been accounted for (Martin et al., 2021).Cooperation can also help local stakeholders to innovatively redeploy resources after a crisis, improving local recoverability.
Partly for such reasons, cooperation is a core element of place-based policies, whose 'ultimate purpose' is 'to induce private agents, individually and through "voluntary institutions", and local governments to do what they fail to do by themselves, which is investing time and effort in revealing knowledge and preferences and aggregating them so as to provide public goods and services' (Barca, 2009: 45).Cooperation is necessary both for the articulation of local visions of development and for the design and implementation of tailored strategies to implement them.The literature on place-based policies emphasises the importance of building local institutional capacities for dialogue and collective decision-making, which are often lacking, particularly in left-behind places situated in highly centralised national contexts (Beer et al., 2020: 42;Martin et al., 2021: 118).Overcoming the obstacles to local cooperation is key to the success of such local institutional arrangements.
Whether it occurs spontaneously or in the framework of place-based policies, local cooperation can generate benefits not only for local productivity but also in terms of fostering a stronger local identity, which can positively affect people's experience of place and community.This can help address the emotional side of the left-behind problem, which concerns the disappearance of a local sense of purpose that comes with the decline of traditional industries (Telford, 2022).
Despite the centrality of local cooperation for achieving the goals of place-based policies, there are still important gaps in our understanding of how to trigger it, particularly in disadvantaged areas that experienced an erosion in their social capital.As argued in the 2009 Barca report, which outlined a place-based approach to EU Cohesion Policy, 'the mobilisation of local actors, the opening up of opportunities for innovators and the limitation of rentseeking have not been helped by a lack of reference to, and debate on, the policy concept' for achieving those goals (Barca, 2009: 101).The Barca report put forward several proposals to facilitate local stakeholders to 'experiment with solutions while exercising mutual monitoring' and to create incentives 'for local actors to reveal information, to risk and to invest' (Barca, 2009: 177).These proposals concern the concentration of Cohesion Policy resources on fewer priorities, the sharper definition of objectives, an improved system of indicators and targets and additional efforts to raise local awareness.Nevertheless, it is not clear if these proposals are sufficient for overcoming the obstacles to cooperation in adverse settings (Barca, 2009: 177-178).More generally, there remains 'a notable absence of clarity for public sector authorities charged with mapping out new place-based frameworks', particularly regarding 'the circumstances and structures that provide fruitful ground for place-based policy' and 'the role of governance, and innovation in governance, as a pathway to the successful implementation of place-based policy' (Beer et al., 2020: 15).
To better understand how to promote local cooperation in adverse settings, it is important to be able to diagnose the nature of the obstacles that inhibit it.Although it is widely recognised that local cooperation is hard to achieve, especially in disadvantaged areas, there are two broad ways to approach the question of what precisely makes it so difficult.
On the one hand, political economists writing from the perspective of rational-choice institutionalism often assume that local actors have sufficient information and cognitive capacity to engage in best-response maximization, that is, that they operate in a framework of substantive rationality.Game theory provides useful tools for understanding the obstacles to cooperation that actors face when the range of available strategies and the associated benefits and costs are relatively clear to them, as is often the case in economic exchanges.Thus, political economists often analyse the factors that inhibit cooperation in terms of collective action problems, where 'the individual pursuit of self-interest generates socially undesirable outcomes' (Ferguson, 2013: 4).
The structure of the collective action problems that local stakeholders face will depend on the sector and type of cooperative activity in question.Vertical cooperation along the supply chain is susceptible to hold-up problems, which arise when one party to an exchange is required to make investments in specific assets that later put her in a position of vulnerability to be strategically held up, that is, to be stripped of part or all of the benefits of those early investments, by an opportunistic other party (Ferguson, 2013;Lorenz, 1988).Hold-up problems frequently inhibit the implementation of quality improvements that require costly decisions to be taken upstream in the supply chain based on promises of rewards that are removed in time.
In turn, horizontal inter-firm cooperation is susceptible to problems arising from the disinclination of local actors to bear the costs of strategies that have strong positive externalities.If each player can only capture a small share of the benefit from contributing to a collective good, the dominant strategy will be not to contribute.This situation can be captured by a Prisoner's Dilemma game.As famously pointed out by Olson (1965), in large groups, individual contributions to a good shared among all group members are so small that they are barely noticeable.As a result, individuals have the incentive to free-ride on others' contributions, leading to underinvestment, or even to the non-supply of the collective good.
Local actors also face distributional conflicts, which arise when multiple cooperative strategies are available, but the choice of cooperative strategy affects the distribution of the payoffs from cooperation.This problem can be illustrated with a battle-of-the-sexes game, where both players have an interest in adopting the same cooperative strategy, but if they both insist on coordinating on their preferred strategy, cooperation may break down altogether.
While these game theoretical concepts underpin how many political economists approach cooperation, sociologists challenge the underlying assumption of substantive rationality.Sociological neo-institutionalism emphasises the centrality of 'cognitive scripts, categories and models' (Hall and Taylor, 1996: 948) in shaping actors' preferences.It takes an expansive view of the range of goals that individuals may seek to pursue in the process of expressing their identity, but its account of the influence of cognition is relevant to the study of cooperation even under this article's more restrictive assumption that local stakeholders usually seek to improve their material Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 well-being (for a similar approach, see Culpepper, 2003).If 'shared cognitive frameworks ( … ) inform agents about how the world works and what is in the agents' interest' (Emmenegger, 2021), then the reproduction of uncooperative local institutional equilibria may be due not so much to collective action problems, but more fundamentally to individuals not being in a position to 'even conceive of appropriate alternatives (or because they regard as unrealistic the alternatives they can imagine)' (DiMaggio and Powel, 1991: 11).
In line with this type of argument, local stakeholders often lack the cognitive tools required to understand the potential benefits of cooperation and even to grasp the nature of available cooperative strategies: they are 'unclear about what game they are playing and consequently how they should understand their own stakes in the game' (Culpepper, 2003: 16).Substantive rationality models cannot accommodate these kinds of high cognitive obstacles: instead, a bounded rationality approach must be used in such cases.Boundedly rational actors remain goaloriented, but rather than selecting the utility-maximizing response, they make decisions relying on heuristics, or 'mental procedures that readily combine various inputs from current and prior experience to produce impressionistic judgements' (Ferguson, 2013: 126).To trigger cooperation, the main challenge is to alter the mental models used by local actors to make decisions.
One of the most important cognitive obstacles to cooperation concerns entrepreneurial discovery.People typically have major knowledge gaps regarding the goods and services demanded in far-away markets (Shane, 2000).Habituated to specific production methods, local firms and other stakeholders may fail to conceive of alternative, cooperative strategies.Given that the short-term costs of switching to a new strategy are easier to calculate than the fundamentally uncertain future benefits, local actors may also fail to correctly estimate the expected payoffs of alternative cooperative strategies (Ostrom, 1990).Entrepreneurial discovery is hence an important challenge for cooperative efforts.The relative importance of this obstacle will be higher, the more innovative the form of cooperation in question.
Scholars of social networks and participatory governance point to social fragmentation as another reason why boundedly rational actors may fail to correctly assess the benefits and costs of cooperative strategies.Social fragmentation can generate subjective perceptions of unbridgeable differences even when objectively, cooperation could pay off.Members of particular social groups may be oblivious to their shared interests with other groups: it is only through a process of 'coming to a common, and generally surprising view of an economic situation which each thought it had understood fully, [that] mutually suspicious groups can redefine their relations and (prudently) begin to construct communities of interest ( … ) where none had seemed possible' (Sabel, 1993(Sabel, : 1149)).In this view, the fundamental obstacle to cooperation is fragmentation among distinct groups of local stakeholders (Granovetter, 1973;McDermott, 2007;Streeten, 2002).
While the political economists' and sociologists' accounts of the obstacles to cooperation are sometimes presented as incompatible with each other, both approaches are useful for understanding the types of obstacles that economic actors face in the field.Collective action problems and cognitive obstacles to cooperation both arise in real-world settings, but they do so sequentially: the resolution of cognitive obstacles to cooperation must occur prior to the incidence and resolution of collective action problems.Until local actors acquire a basic understanding of the different strategies that are available to them and of the payoffs associated with each, collective action problems remain inert.However, once the actors start operating within a problem-complexity boundary where the costs and benefits associated with different strategies can at least be estimated, they are still likely to face conflicts of interest and collective action problems (Ferguson, 2013), which a purely sociological approach may miss (Hall and Taylor, 1996).Moreover, the cognitive obstacles to cooperation can be expected to be bigger, the higher the degree of innovation required in the specific sector and type of cooperative activity in question.On the other hand, when it comes to more incremental types of innovation, the payoffs of different strategies will be clearer to the actors involved and collective action problems will act as direct and immediate obstacles to cooperation.This analytical framework is consistent with a 'broad conception of rational action' (Ostrom, 1990: 37), in which economic actors take decisions based on a calculation of the costs and benefits of different strategies, but at any moment in time may face high degrees of uncertainty about the nature of those costs and benefits.
Using the sequential, integrated theoretical framework proposed here to diagnose the obstacles to cooperation in left-behind places allows us to draw on the cooperation literature in political economy and sociology to design appropriate solutions, tailored to the specific obstacles to cooperation observed in each context.These solutions can be provided at different levels.At the local level, the literature on place-based policies emphasises the important role of 'erudite and charismatic local leaders' (Martin et al., 2021: 11), whose concrete strategies for mobilising other local actors can include, for example, the incremental construction of trust to overcome hold-up problems (Farrell, 2009;Lorenz, 1988), or the gradual dissemination of a novel entrepreneurial vision, based on a few demonstrable, early successes, to overcome cognitive obstacles to cooperation (Gartzou-Katsouyanni, 2022).On the other hand, macro-level policies can be designed to provide positive or negative selective incentives that can reduce free-rider problems (Olson, 1965).Alternatively, they can Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 incentivise diverse local stakeholders to 'study their industries jointly' (Sabel, 1993(Sabel, : 1158(Sabel, -1159) ) and deliberate 'about shared problems, how to interpret them, and how to solve them' in inclusive arenas for discussion, creating 'shared cognitive maps' and thereby addressing the problem of social fragmentation (Hall and Taylor, 1996: 950).The suitability of these and other local and macroinstitutional strategies will depend on the analysis of the most important obstacles that apply in each area.Thus, carefully combining insights from rational-choice and sociological neo-institutionalism in explaining processes of institutional change can yield analytical benefits as well as practical implications for policy design (Emmenegger, 2021;Hall and Taylor, 1996).

Empirical setting and methods
The utility of the theoretical framework outlined in the previous section will be illustrated based on empirical evidence from four pairs of matching case studies within an overall setting that is unfavourable to cooperation, namely Greece.Greece is a good example of a context with neither the cultural propensity towards trust-based relations that underpins successful inter-firm cooperation in the Third Italy (Becattini et al., 2009;Musotti, 2009;Putnam, 1993) nor the robust framework of formal institutions that support similar forms of cooperation in places like Baden-Württemberg (Farrell, 2009;North, 1990).The country scores low on surveys that measure interpersonal trust: in a recent survey, the share of respondents who agreed that most people can be trusted was 24% in Greece, compared to an EU average of 47%, 58% in Germany and 85% in Finland (European Commission, 2018).Greek formal institutions also have well-documented deficiencies that make cooperation harder.The hyper-centralisation of Greece's political and administrative system decreases local stakeholders' opportunities for deliberation and decision-making (Loughlin, 2001), while administrative and judicial weaknesses undermine the uniform implementation of the law (Doxiadis, 2014).The literature on the governance of local economic systems, including work on the smart specialisation paradigm, is 'often informed by the experience of more prosperous regions' (MacKinnon et al., 2022: 42), which specialise in 'advanced, knowledge-rich, high-tech sectors' (Burroni et al., 2008: 474).Complementing this literature with evidence from less favourable contexts can yield useful insights for leftbehind places, which often also face an erosion in social capital and a low 'strategic policymaking and delivery capacity' at the local level (Martin et al., 2021: 101).
Despite the adverse contextual conditions, cooperation is important for the future of the Greek economy, which relies heavily on SMEs 1 (Doxiadis, 2014).Greek regions were particularly likely to face a development trap over 2001-2015 (Diemer et al., 2022).This means that many Greek areas found it difficult to retain their past economic dynamism and to perform on par with their European peers during this period.The stakes rose sharply with the Eurozone crisis, which brought about the collapse of the country's traditional, inward-oriented growth model based on big public spending.Local stakeholders would have to switch to alternative, export-oriented methods of production to retain their area's prosperity.While many places continue to struggle, several examples of cooperative efforts among local firms and other stakeholders to achieve export-oriented growth can be identified in Greece in recent decades.Some of these efforts 'have even managed to revitalise whole areas with their success' (Vakoufaris et al., 2007: 779).These experiences are relevant for other leftbehind places that have seen their traditional economic models decline and that face the challenge of improving their economic performance through cooperation.
The following two sections analyse the obstacles to cooperation in four pairs of case studies of specific sectors in specific areas of Greece.Each pair includes a case where specific types of cooperation emerged within the last 40 years, contributing to the economic vitality of a previously struggling location, and another, similar case, where cooperation could have had comparable benefits, but it did not arise.The cases where cooperation successfully emerged provide a unique opportunity to study the obstacles that had to be overcome for local stakeholders to productively work together in adverse circumstances, while their matching pairs offer an indication about whether similar obstacles continue to hinder cooperation elsewhere.The cases are drawn from the agri-food and tourism sectors, responding to calls to focus attention on 'the "long tail" of lower value-added sectors' that play an important role in several left-behind areas around the world (MacKinnon et al., 2022).
The sub-sectors that the four case study pairs represent were selected to make it possible to examine whether the relative importance of different types of obstacles to cooperation varies in the way that the proposed theoretical framework suggests.The collective action problems that stakeholders are expected to face are different in the agri-food sector, where vertical cooperation along the supply chain is particularly important for quality upgrading (Chappuis and Sans, 1999), and in the tourism sector, where horizontal cooperation can influence the characteristics that make a place attractive to visitors (Gordon and Goodall, 2000).Moreover, the relative magnitude of the cognitive obstacles to cooperation is expected to vary depending on the degree of innovation required in each case.To explore this relation, two sub-sectors were chosen within each of the agri-food and tourism industries, including an established sub-sector, where the main challenge was to upgrade quality, and a non-established sub-sector, where a higher degree of innovation was needed.This method of systematic comparisons among a medium-n number of Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 cases allowed for the study of mechanisms using qualitative evidence, on the one hand, and for abstraction based on the analytically relevant features of more than one or two very particular cases, on the other (Bates et al., 1998;Bennett & Elman, 2006).Table 1 summarises the classification of the case studies, while Figure 1 shows their location.
The first pair of case studies focuses on the wine sector of the islands of Santorini and Lemnos.Santorini has been transformed from an area producing low-quality, bulk wine that was 'useless' for the tastes of modern consumers in the 1980s (interview #13) to 'a "flagship" wineproducing region that is leading the way for building awareness of Greek wine not only in the US, but also worldwide' (Kotseridis et al., 2015: 43).Cooperation among pro-ducers and winemakers in the areas of quality upgrading and marketing has contributed to that result (Iliopoulos and Theodorakopoulou, 2014;Vlahos et al., 2016). 2 The smaller extent of cooperative activities in Lemnos's similarly structured wine sector is associated with worse economic outcomes, including much lower producer prices, which translate into lower producer income even if one takes into account that yields in Lemnos are higher (Table 2).It is also remarkable that despite having 5.4 times more utilised agricultural land than Santorini, Lemnos only has 53% of Santorini's turnover in food processing and manufacture overall and 36% of Santorini's turnover in the manufacture of beverages specifically.Moreover, even though the booming tourism sector offers lucrative opportunities to convert agricultural land to accommodation Moving on to the established part of the tourism industry, the second case study pair involves mass tourism in Santorini and Chalkidiki.Inter-firm cooperation to upgrade quality can be necessary for successful touristic destinations to survive over time (Morgan, 1991).In Santorini, concerted efforts to upgrade quality, combined with the fact that local stakeholders were among the first in Greece to attract conference, wedding and gastronomical tourism, have contributed to the island having one of the longest tourism seasons in Greece and benefiting from substantially higher per capita spending by tourists than the country average (€174 per person per night in 2016 compared to €69 in Greece overall) (Spilanis, 2017: 23, 45).In contrast, stakeholders in Chalkidiki, one of Greece's earliest and biggest mass tourism destinations, strongly feel that the failure to upgrade quality and lengthen the season is affecting them adversely (interviews #21,  2).The third pair of case studies zooms into two unique agri-food products, Chios mastiha, which is a type of resin gathered on Chios island, and Kozani saffron.In the postwar period, these products progressively lost their price competitiveness in international markets, while they traditionally had next to no domestic uses.As a result, creating new markets for innovative goods utilising those local inputs was central to the resilience and growth of both local sectors.Producers and processing firms were more successful in achieving this goal in Chios than in Kozani, and cooperation to create a range of new mastiha products, to market them in Greece and abroad through a new network of MastihaShops since 2002, and to upgrade the quality of raw mastiha played a key role.Compared to the vibrant firm ecosystem around Chios's mastiha cooperative today, the number of private firms that Kozani's saffron cooperative collaborates with to produce saffron products is very small.Mastiha production has increased from 83 tons in 2003 to 216 tons in 2021 (a 40-year record).The producer's price also rose from 49,16€/kg in 2001 to 74,12€/kg in 2009 and 85-90€/kg in 2021-2022 (Agrotipos, 2022;Lioukas, 2013).While comparable data is not publicly available for Kozani's saffron, during fieldwork a local producer claimed that 'in the last 3-4 years, the price has been falling' (interview #54), and there was widespread concern about the gradual lifting of the embargo on Iran, which is the world's biggest saffron producer.Employment in the primary sector has been falling more slowly in Chios (21% decline between 2001 and 2011) than in Kozani (50% decline) (Table 2). 3  Finally, the fourth case study pair focuses on alternative tourism in areas threatened by depopulation.It juxtaposes the experience of the remote, mountainous village of Nymphaio, which was revitalised through a set of local initiatives to attract visitors and restore buildings following the traditional architectural style, and that of Ambelakia, which did not witness equivalent cooperative efforts despite being better endowed in terms of accessibility and cultural heritage to develop as an alternative tourism destination.Between 1991 and 2001, during the peak period of these cooperative activities in Nymphaio, the village's population rose by 69%, while the population of Ambelakia declined by 8% (Table 2).During the same period, the population of the country's 1337 rural mountainous villages having 100-500 residents in the 1991 census rose on average by 5%, and the population of the country as a whole rose by 7%.Moreover, in the 2011 buildings census, Nymphaio had double the number of hotels than Ambelakia.
Information about the potential benefits of cooperation to local stakeholders, the extent of their attempts to engage in cooperative activities, and the obstacles they encountered in that process was collected through a variety of sources.Those included 86 semi-structured interviews with producers and representatives of cooperatives, firms, local authorities and civil society groups, depending on the sector (Appendix A), as well as documentary sources, the local press and grey literature about the case study areas (Appendix B).

Upgrading in the wine sector: hold-up problems and distributional conflicts in lowtrust, competitive contexts
In the 1980s, many of the wines that were produced in Greece were heavy and unsuited to the tastes of the modern consumer.Several wine cooperatives offered higher than market prices to grape producers and accumulated debt, which they serviced with loans from the Agricultural Bank of Greece (Efthymiou, 2017).This model of development ended abruptly in the early 2010s with the collapse of the Agricultural Bank, which led to many cooperatives facing severe financial problems and even bankruptcy.
Santorini's wine sector avoided this trajectory, and two forms of cooperation, linked to quality upgrading and marketing, respectively, contributed crucially to that outcome.First, starting in the late 1980s, winemakers and grape producers engaged in vertical cooperation along the supply chain and began upgrading the quality of the local wine.Yannis Boutaris, a northern Greek winemaker who had just started producing wines in Santorini, considered that one of the first changes needed for producing lighter, more modern wines was to shift the timing of the harvest of grapes from September to August.Producers greeted the change 'with much suspicion': early-harvested grapes would weigh less, 'so the producers thought that Boutaris is moving up the harvest in order to pay for fewer kilos and give them less money' (interview #30).A context of widespread mistrust towards private enterprise made it harder to overcome those concerns.As a producers' representative wrote to the wine cooperative's General Assembly in 1989, 'let all the producers understand that Mr. Boutaris came to Santorini so that he can make money while they work': if Boutaris was not willing to pay 'a price that is in the interest of the producers' for obtaining local grapes, 'he should close down his factory and leave' (Santo Wines General Assembly meetings, Act 117, July and August 1989).It took years for Boutaris, in cooperation with the cooperative's management board, to gradually establish his credibility and introduce changes in cultivation and harvesting methods that were increasingly demanding in terms of their upfront costs.
The second form of cooperation in Santorini's wine sector was a joint marketing campaign in the 2010s called Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 'Wines of Santorini', which targeted the North American market.Santorini's wine cooperative worked together with the majority of the island's 15-18 private wineries in this project, which helped the local sector overcome the collapse of domestic demand for luxury wines during the crisis.The main obstacle to this effort was distributional conflict, as different project designs would benefit some types of wineries more than others.From the perspective of a local winemaker, 'there are two tiers of Santorini wines.If the project prioritises the sale of second-tier wines, this doesn't help us so much, because the importers cannot absorb our bottles by raising the prices enough' (interview #34).These problems were exacerbated by a context of high competition among the wineries for the increasingly expensive local grapes: 'we have an issue with the raw material, the grapes are limited.This creates conflicts, and sometimes those conflicts create siloes' (interview #71).Overcoming distributional conflicts in this competitive environment required leadership by the wine cooperative and a US wine importer, as well as a funding programme by the EU's Common Agricultural Policy for collaborative promotion activities to third countries.
The winemakers in Lemnos were fully aware that a joint marketing effort was in their collective long-term interest: 'the most important thing for the sector is to create lovers of Lemnos' (interview #85).Nevertheless, similar distributional conflicts as in Santorini prevented them from engaging in cooperation: in a context of intense competition for a share of Greece's crisis-ridden domestic market for wine, 'perhaps it's a bit premature to go as five competitors to an exhibition and knock on the same door, and say come, choose the best, we don't mind' (interview #85).On the other hand, vertical cooperation along the supply chain is easier in Lemnos today than it was in Santorini in the 1980s.The inability of the local cooperative to freely borrow money means that local grape producers have fewer outside options: they cannot simply tell the private winemakers to pay a high price or shut their wineries and leave.Thus, the local winemakers claim that 'it's extremely easy' today to find grapes at the specifications they require: 'it's the producers who are begging to bring us grapes' (interview #83).While this situation reduces the obstacles to quality upgrading, it also creates the potential for skewed distributional outcomes against the local producers.
Thus, the biggest obstacles to cooperation in an established agri-food sector like wine can be analysed as collective action problems, which are aggravated by specific contextual conditions.Vertical cooperation along the supply chain is inhibited by hold-up problems, where agricultural producers are reluctant to pay the short-term cost of implementing even minor changes, as they doubt that the promised future benefits will materialise.These problems are harder to overcome when trust is low and the bargaining power of upstream actors is high.In turn, hori-zontal cooperation among a small number of wineries is hindered by distributional conflicts, which are particularly difficult to resolve when competition is high.Boosting cooperation in the sector requires tailored solutions that address these problems.

Upgrading in mass tourism: the management of horizontal externalities in weak institutional settings
The structure of the mass tourism sector differs from that of the wine sector.A few thousand small businesses operate in mass tourism destinations with fragmented ownership structures such as Santorini and Chalkidiki.Successfully managing those firms' horizontal externalities can lead to quality upgrading and a longer tourism season (Gordon and Goodall, 2000;Healy, 1994), which has to some degree happened in Santorini, especially in the villages of Oia and Imerovigli, but remains an aspiration in Chalkidiki.Failing to manage tourism firms' horizontal externalities can generate overuse and underinvestment in the common resources that make a destination attractive, which is a problem both in Chalkidiki and in parts of Santorini.
Coordination problems tend to pose some of the biggest obstacles to cooperation for quality upgrading.Local firms are reluctant to take the decision to upgrade individually, as they will not be able to reap the full benefits of upgrading unless a critical mass of other firms also upgrade at the same time.Coordination problems remain a powerful impediment to upgrading quality and lengthening the tourism season in Chalkidiki, even though local stakeholders see both aims as crucial for their prosperity.One interviewee explained: 'retail shop owners have inundated the market with cheap Chinese products.But they cannot do otherwise because the clients don't have money' (interview #22).'Tourism of a very low quality has developed', and hotels, restaurants and shops all contribute to the problem: 'everyone must improve quality for the level of the clients to rise' (interview #25).Santorini's experience shows that the tourism entrepreneurs who went ahead and upgraded before a critical mass of other firms did so, paid a disproportionate cost while only reaping a fraction of the benefit.In the account of the founder of one of the first restaurants on the island to offer upscale Greek cuisine utilising local ingredients, Giorgos Chatzigiannakis: "When the local wine became lighter and more refined, we had to find a way to utilise local products in a different way.We worked very hard with colleagues, producers, winemakers.We invited journalists.And now, 650 restaurants operate on this little island -imagine!(…) A journalist once asked me, what have you achieved in Santorini, Mr. Chatzigiannakis?And I said, well, I've achieved that everyone is cooking now!" (interview #35).
Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 Local leading actors such as Chatzigiannakis, who are willing to undertake the risk and cost associated with upgrading first, can help overcome coordination problems in mass tourism destinations.Nevertheless, stakeholders in both Santorini and Chalkidiki have found that purely local solutions rarely suffice to address overtourism, which is another problem in both destinations (Gounaris, 2015;Spilanis, 2017).Limiting overtourism structurally resembles the problem of managing a common-pool resource (Blanco, 2011), and it tends to stumble on a particularly challenging combination of free-rider and distributional problems.Most stakeholder groups in Santorini agree that 'overtourism is a big problem on the island' (interview #75), but each group prefers to pass the burden of adjustment to others.Tellingly, an incumbent hotel owner considered that 'the building of new hotels should not be permitted' (interview #75).Α constructor of pools agreed about the problem but disagreed about the solution: 'When everyone has done their own thing, you can't just punish the last guy to come along' (Tsiros, 2018: 28).Despite having commissioned several studies which recommended strategies to limit overtourism (EBRD, 2019;Spilanis, 2017), the local authorities have so far failed to spearhead the adoption of a comprehensive solution.As is known from Ostrom's (1990) work, a macro-level actor, such as a court or a national government, which threatens to impose its own cutbacks if the locals fail to take action to limit the overuse of a common-pool resource, can succeed in mobilising local solutions to the problem.Nevertheless, Greek macro-level institutions do not provide such incentives, while the growth of an exceptionally underregulated sharing economy sector has made it hard to even identify the relevant group of industry stakeholders who would need to reach a local agreement.
Overall, collective action problems that relate to the management of horizontal externalities among larger groups of actors constitute a powerful tool for analysing the obstacles to cooperation in mass tourism destinations.Such analysis can help design local strategies and macrolevel policies that address the specific incentive structures that inhibit cooperation in the types of places in question.

Creating innovative products in the mastiha and saffron sectors: Entrepreneurial discovery in a statist environment
Local stakeholders in case studies of established subsectors within different industries tend to understand the potential benefits of cooperation, even though they are often inhibited from reaping them by collective action problems.This is frequently not true in non-established sub-sectors, where cooperative activities entail a higher degree of innovation.In such cases, cognitive obstacles to cooperation must be overcome before collective action problems even arise.
Chios mastiha and Kozani saffron are both unique agricultural products that were traditionally exported in raw form to international clients.Mastiha was traditionally exported to Turkey and Iran, but by the end of the 1970s both countries switched to cheaper substitutes, creating a crisis in the sector (Tsouhlis, 2011).In turn, Kozani's saffron exports are vulnerable to competition from Iran, which offers prices as much as ten times cheaper than those in Kozani (Kozani saffron cooperative, 2001;Palaiologos, 2016).Under circumstances of sanction-free economic exchange with Iran, 'it was easier to sell stones than to sell saffron' (interview #53).Producing and marketing innovative, differentiated final products using Chios mastiha or Kozani saffron can address the vulnerability of both sectors to the vicissitudes of international markets.This strategy requires cooperation, as local producers and agri-food firms are too small to achieve those goals individually.
Discovering this strategy was difficult as neither mastiha nor saffron were traditionally used locally.When the Chios cooperative founded a subsidiary company to sell mastiha products in 2002, 'they had to look under the stones to find mastiha products' (interview #59).According to Emmanouil Patsilias, a former President of the saffron cooperative in Kozani, his predecessor had refused to conduct research in the health properties of saffron, saying: 'what if it turns out that the product is harmful for health?Do you want us to be ruined?'Patsilias commented: 'They thought that some naïve people were buying saffron because they wanted to damage their health.These things had to be overcome, they were ridiculous' (interview #53).
This hurdle is particularly difficult to overcome in places that traditionally followed statist models of economic development.In such areas, local actors are habituated to thinking that economic prosperity comes through interactions with the state, rather than through entrepreneurship and innovation.A strong statist orientation was one of the defining characteristics of Greek agricultural cooperatives.Not only was the mastiha cooperative financially rescued by the state several times (Tsouhlis, 2011), but the producers were also accustomed to thinking of the cooperative as 'our home: when we didn't have money, we would go and take out a loan' (interview #61).In this environment, it was nothing short of radical to think that a cooperative was not a 'social organisation' supposed to support producers through loans, cheap groceries and administrative jobs for their children, but it was 'a commercial business with the sole aim of making profit' (interview #63).A statist orientation is embedded in Kozani also because the area's recent development has been highly reliant on lignite mining by the state electricity company.It is indicative that most discussion about economic development in the local press concerned either Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 the future of the local lignite reserves or the distribution among social groups of the resources from the so-called 'local fund', which is given by the state as compensation for the pollution generated by the mines (for example, see the 16/4/2000 issue of the Kiriakatikos Chronos newspaper).Overcoming state-oriented mental models and devising innovative, collective entrepreneurial strategies is not an easy task from this starting point.
Social fragmentation among distinct groups of stakeholders within a local sector can exacerbate the obstacle of entrepreneurial discovery.In Kozani, the saffron farmers' cooperative and the local agri-food entrepreneurs perceive the distance that separates them to be so big, that mutually beneficial strategies seem almost inconceivable.As mentioned by an entrepreneur, 'the problem [with the saffron cooperative] is that there is no openness, there are no Presidents with an open mind.So I seek personal benefits, I can't open everyone's eyes and lose out myself' (interview #52).In turn, a representative of the saffron cooperative's management told me that there is cooperation with one or two local firms, 'but these links are very limited, and we don't want them to expand' (interview #55).Similarly, when the Chios mastiha cooperative started producing its own final goods, 'we met rabid reactions.It's enough to look at the press.( … ) Businesspeople came out saying that "your role is to cultivate the fields, we are the ones who sell", they took us to court' (interview #63).
Despite the adverse circumstances, the obstacles of entrepreneurial discovery and social fragmentation were overcome in Chios.This was largely thanks to the leadership of the mastiha cooperative, which followed a strategy of incrementally proving the benefits of its innovative entrepreneurial strategy and taking costly actions to establish a conception of shared interest with private firms.Once the cognitive obstacles to cooperation were addressed, hold-up problems and distributional conflicts also arose there.Indicatively, 'very many things have changed regarding quality' in the production of mastiha.These changes made 'the life of the producers harder' and entailed a financial cost (interview #58).Thus, the cooperative had to resolve collective action problems that arise during vertical cooperation along the supply chain.Nevertheless, when a high degree of innovation is required for cooperation to succeed in a local sector, the first obstacles that need to be addressed are cognitive.This is particularly important in contexts where the collaborative entrepreneurial strategies in question are dissonant with traditional local production models.

Creating an alternative tourism destination: adopting collective entrepreneurial strategies in a context of depopulation
Nymphaio and Ambelakia are both villages that thrived during the Ottoman period, but experienced economic decline in the twentieth century and faced accelerated de-population in the postwar period.For Nymphaio, the final blow came with the development of well-paid jobs in fur trade in a nearby town in the 1960s and 1970s: 'Before the development of fur in Kastoria we had 1000 residents, whereas afterwards only 50 were left' (interview #15).Similarly, in Ambelakia, 'many residents left as workers in [the nearby city of] Larissa, and they only came back during the weekend.Those who remained tried to survive with animal husbandry and agriculture: wines, chestnuts, small cultivations' (interview #48).
In such contexts, envisaging that young and wealthy urbanites might wish to spend their holidays in the village, and conceptualising what types of activities could attract them there, are far from trivial challenges.These challenges pose different kinds of obstacles to cooperation than the obstacles faced in established tourism destinations.As it happened, in Nymphaio, one of the core attractions that was created to attract visitors was a sanctuary for brown bears by the NGO Arkturos.The former President of the Nymphaio Community recalls that initially, the village residents found it difficult to imagine that such a sanctuary could attract visitors: 'at the start, the locals were mocking Arkturos, saying "hear, hear, the bears will come to feed us!"But later they understood.Arkturos, which we mocked at the start, proved to be a goldmine for the village' (interview #14).
In Ambelakia, these challenges are compounded by a widespread belief that the state is responsible for resolving the village's problems.This belief has to do with the symbolic importance of Ambelakia in the Greek collective imaginary as the place where 'the first cooperative in the world' emerged in the eighteenth century.Preserving this cultural heritage is seen as a national duty.As the President of the local Folk Art and Historical Museum complained: 'The state has not showcased the history of the Ambelakia cooperative like it should have.In the past this was the responsibility of the prefecture authorities, now it's the responsibility of the periphery-level government.Finally, they should come and see the [local attractions] and preserve them' (opening remarks at a public event, Ambelakia, 15/7/2018).
The need to involve a broad range of stakeholders in creating tourism attractions makes such efforts vulnerable also to social fragmentation.In both Nymphaio and Ambelakia there was a tendency for mutual perceptions of unbridgeable differences between locals and outsiders.Ambelakia has several local associations, but they intensely mistrust each other.As mentioned by a member of one of the village's associations, '[Association X] is a step for the political ascent of its members.Whereas we don't do politics, we want the village to develop'.Reciprocally, according to a member of Association X, 'the question about Association Y is whether they do what they do in order to attract tourists, or whether they do it to bring in their own acquaintances and friends' (interviews #50, Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 #51).Underlying this conflict was an antagonism about who was most genuinely an insider: 'Here in the village they have this mentality, that outsiders like us want to show off.( … ) But those at [Association X, who originate from Ambelakia], are not concerned with earning an income here, as most live in Larisa anyway' (interview #50).Similarly, in Nymphaio, it was thought that one of the reasons for the locals' original resistance against the establishment of the bear sanctuary was that 'they had a complex, that these guys will come to take over our village' (interview #13).
Thus, overcoming the obstacles of entrepreneurial discovery and social fragmentation appears important in non-established sectors more generally, regardless of the industry.Once those cognitive obstacles were overcome, collective action problems also arose in Nymphaio.Given the considerable levels of investment required to create new tourism attractions, many of those issues had the structure of a public-good provision problem.
Table 3 summarises the main initial obstacles to cooperation encountered in each of the four pairs of case studies, as well as the contextual conditions that aggravated those obstacles.

Conclusion
Triggering local cooperation is a key aim of place-based policies for left-behind areas, as it can help improve economic performance and restore a local sense of agency and community.Nevertheless, it is not always clear 'how diverse groups of actors [can be] encouraged to work towards a common purpose, [and] how their actions can be coordinated for the collective good' (Beer et al., 2020: 48).To address those questions, it is important to firstly understand precisely which factors tend to hinder cooperation in different types of local contexts.
Combining insights from the political economy and sociology literatures on cooperation, this article has put forward an integrated theoretical framework that can be used to diagnose the nature of the obstacles to local cooperation, depending on the sectoral structure and challenges faced in each place.It was argued that in established sub-sectors, local stakeholders often know that cooperation can generate collective benefits.However, cooperative efforts tend to stumble on collective action problems, whose nature depends on the structure of the sub-sector in question.On the other hand, when it comes to cooperation for engaging in more innovative types of economic activity, a series of cognitive obstacles to cooperation must be resolved before collective action problems even arise.The utility of the proposed framework was illustrated with a medium-n comparative case study approach.This research design allowed for the use of qualitative evidence to study the concrete obstacles to cooperation in specific settings, whose structure varied in analytically relevant ways.
The proposed theoretical framework opens the way for utilising a set of powerful analytical tools from the political economy and sociology literatures to design tailored and effective solutions to the obstacles to cooperation in different locations.At the local level, shifting from an uncooperative to a cooperative equilibrium often requires 'skilled action' by a small group of local leaders (Emmenegger, 2021: 620).Their tactics can include, among others, the incremental construction of trust, as in Santorini's wine sector (Farrell, 2009); shouldering the first-mover disadvantage of horizontal quality upgrading efforts, as in Santorini's tourism sector (Ferguson, 2013); showing in practice that a particular sub-group of local actors can be useful allies to others, as in Chios's mastiha sector; and incrementally demonstrating the viability of a novel entrepreneurial idea, as in Nymphaio's alternative tourism sector.While policymakers working on regional development cannot supply left-behind areas with successful local leaders, they can ask local stakeholders to identify the obstacles to cooperation in their area and work with them to design tailored solutions.Policymakers in regional development and other policy areas, such as education, can also promote background conditions that increase the likelihood of local leadership emerging, for example by encouraging locals to participate in translocal networks, which can inspire economic and institutional innovation (Crouch, 2005).
Greater precision in diagnosing the obstacles to local cooperation can also guide the adoption of macro-level institutional frameworks that make it easier for local stakeholders to engage in cooperative activities (Gartzou-Katsouyanni, 2020).On the one hand, facilitative macroinstitutions can encourage local stakeholders to adopt rules that address their collective action problems.For example, in the mass tourism sector, the state can threaten to impose limitations to the overuse of common resources centrally if local actors fail to do so on their own.On the other hand, facilitative macro-institutions can also help address the cognitive obstacles to cooperation by providing opportunities for deliberation and by subsidising the upfront costs of cooperation, which are the hardest to cover locally.Facilitative macro-level institutional frameworks can be provided both within the scope of regional development policy and by sectoral policies addressed to the agri-food sector, tourism, or other sectors.
The article's empirical findings are drawn from a national setting in which local stakeholders in several areas faced challenges that are also relevant for leftbehind places in other countries.In a low-trust, hypercentralised context, those actors had to come together to switch from an inward-oriented, statist model of development to alternative, export-oriented methods of production.The obstacles to cooperation analysed based on these experiences are directly relevant to leftbehind places with a similar sectoral composition, such as deprived coastal areas in the UK and remote rural peripheries across Europe.Nevertheless, the analytical framework developed in the article can also be used with some modifications stipulated by the sector and context of interest to analyse the obstacles to cooperation in a broader range of left-behind places with many smalland medium-sized firms.While there are no universal recipes for revitalising fragmented, left-behind areas, diagnosing the contextually specific obstacles to cooperation can be an excellent initial step for the design of tailored place-based policies.Further research could explore the required theoretical adaptations for examining the obstacles to cooperation in left-behind areas with more concentrated business ownership structures, where the emphasis would be less on inter-firm cooperation, and more on cooperation between the workers, management and broader community.
Local cooperation alone does not suffice to restore economic dynamism and prosperity in left-behind areas.A comprehensive strategy to address the left-behind problem would involve place-based policies with broader aims, as well as 'greater place sensitivity in national policies and in regulation and mainstream spending' at large (Martin et al., 2021: 102).Nevertheless, in many contexts, the emergence of local cooperation is a key ingredient for breaking away from the left-behind predicament.Therefore, strengthening our understanding of the factors that hinder local cooperation and developing strategies to address them can help design tailored, differentiated and effective place-based policies for a range of communities.
Endnotes 1 The share of persons employed in small and mediumsized enterprises (SMEs) in Greece is 87.9%, compared to an EU average of 66.6%.SMEs produce 63.5% of the total value added in the Greek economy, compared to an EU average of 56.4% (European Commission, '2019 SBA Factsheet: Greece', available at: https://ec.europa.eu/docsroom/documents/38662/attachments/13/ translations/en/renditions/native) 2 For other studies linking cooperation to improved economic outcomes in the agri-food sector, see Bianchi (2001), who notes that inter-firm cooperation for 'product differentiation' and 'the creation of marketing consortia' is key for the continued success of mature agri-food industries (129).Similarly, Lamprinopoulou and Tregear (2011) show that inter-firm networks can 'play a role in improving the marketing performance of SME clusters' (425).3 For Kozani, the geographical unit considered is the municipality of Kozani, where most saffron production is concentrated.These are among the few data points publicly available at the municipality level, which is more appropriate for this case study than the level of the peripheral unit of Kozani (NUTS-3).The peripheral unit is 3.5 times bigger and saffron is only a small fraction of its total agricultural production.

Table 2 .
Background information about the case studies.Sources: Hellenic Statistical Authority (ELSTAT): Annual Agricultural Statistical Survey (area with vineyards; volume of grapes used for wine); Farm Structure Survey (utilized agricultural area); Statistical Business Register (turnover, number of legal units, and employment by economic activity); Population and Housing Census (employment in the primary sector; surface area and altitude of administrative units; de facto population, that is, the number of people present on the day of the census, which was the main population measure used in Censuses until 2001); buildings census (number of buildings exclusively used as hotels).The information about the kilo price of local grapes was collected by the author during fieldwork.Downloaded from https://academic.oup.com/cjres/advance-article/doi/10.1093/cjres/rsad037/7334595 by guest on 20 November 2023 #22, #23).It is telling that the turnover of the hotel and short-stay accommodation sector in Santorini, which has one-tenth of Chalkidiki's surface area and 37% fewer hotels, is 62% bigger than in Chalkidiki (Table