THE CASE FOR REGIONAL DEVELOPMENT INTERVENTION: PLACE-BASED VERSUS PLACE-NEUTRAL APPROACHES†
We are particularly grateful to Michael Storper, Charles van Marrewijk and Frank van Oort, the special issue editors, for their incisive comments to earlier versions of the manuscript. The comments and suggestions by two anonymous referees are also greatly appreciated. Andrés Rodríguez-Pose acknowledges the support of the European Research Council under the European Union's Seventh Framework working Programme (FP7/2007-2013)/ERC grant agreement n° 269868 and of a Leverhulme Trust Major Research Fellowship. The research is also part of the Prociudad-CM programme and of the UK Spatial Economics Research Centre. An earlier version of the paper was presented at a seminar in Rotterdam.
ABSTRACT The paper examines the debates regarding place-neutral versus place-based policies for economic development. The analysis is set in the context of how development policy thinking on the part of both scholars and international organizations has evolved over several decades. Many of the previously accepted arguments have been called into question by the impacts of globalization and a new response to these issues has emerged, a response both to these global changes and also to nonspatial development approaches. The debates are highlighted in the context of a series of major reports recently published on the topic. The cases of the developing world and the European Union are used as examples of how in this changing context development intervention should increasingly focus on efficiency and social inclusion at the expense of an emphasis on territorial convergence and how strategies should consider economic, social, political, and institutional diversity in order to maximize both the local and the aggregate potential for economic development.
Theories of regional economic growth and development have made a quantum leap in the last decades. The seminal contributions of the second half of the 1950s (Solow, 1956; Swan, 1956; Myrdal, 1957; Hirschman, 1958; Rostow, 1959) set the bases of modern growth and economic development theory and for the future implementation of policy in the economic development realm, both at the national and subnational level. They also eventually left the field satisfied with the theoretical solutions, as well as exhausted. Solow's, Myrdal's, Hirschman's, and Rostow's theories dominated thinking in economic growth and development for 30 years, but their impact started to wane as the world in the 1980s and 1990s grew more and more apart from that of the 1950s. As a consequence, scholars started to think about economic growth and development again and this has resulted in two decades of theoretical transformations that have turned the field on its head. First came the endogenous growth revolution (Romer, 1986; Lucas, 1988), the new economic geography (Krugman, 1991, 1995; Fujita et al., 1999) followed, and the institutional turn in economics (Rodrik et al., 2004; Acemoglu and Johnson, 2006a,b) represents the latest twist in this saga.
Yet, and despite this phenomenal recent transformation in theory and empirical analysis, there was one area where little progress had taken place: that of policy implementation. Development intervention at the national and regional level had remained firmly anchored in Solow's, Myrdal's, Hirschman's, and Rostow's thinking, paying little attention, and often no more than lip service, to the recent transformations in the theories of development and growth. Rather narrow concepts, such as the economic base model, have remained the norm in development intervention.
This is all likely to change now as a result of the publication in 2009 and 2010 of a series of highly influential reports about regional development policy intervention by the World Bank (2009), the European Commission (Barca, 2009), the OECD (2009a,b), and the Corporación Andina de Fomento (CAF, 2010), a large Latin American development bank. All these reports on face value deal with similar analytical and policy issues. However, they are inspired by different sources, adopt different philosophies, and arrive at what seem to be very different policy conclusions. If we also include in this list of major works an earlier report (Sapir et al., 2004), then the issues explored by these six reports demonstrate that the debate regarding the case for regional policies is both very much alive and also very complex. At the very least, the debate is far from being settled and requires consideration of a range of fundamental and interrelated issues.
This paper explores the issues raised by these reports and their potential implications for regional policy intervention, drawing on evidence from around the world. The paper is organized as follows. In the next section, we outline some of the major economic geography features of the modern era of globalization and the reasons why a need for a rethinking of regional development has arisen. In Section 3, we examine the rethinking which has taken place, as contained in the above-mentioned reports, and the two quite different schools of thought which have emerged, namely, the space-neutral and the place-based approaches. In Section 4, we consider the differences between the two approaches by discussing two different cases: developing countries and the European Union (EU). As will be seen, the differences center profoundly different understandings of the role played by institutions, governance, and urban hierarchies in influencing historical processes of development. Section 5 then identifies the key challenges and possibilities associated with different approaches to development policy, and Section 6 provides some brief conclusions.
2. GLOBALIZATION AND ECONOMIC GEOGRAPHY
Any discussion today regarding the case for regional policies must first consider the enormous impacts that modern globalization trends are playing in shaping the economic geography of places, irrespective of whether we define places in terms of countries, regions, or cities. The beginning of the modern era of globalization can be traced back to the 1970s (Ferguson et al., 2010), although some of the most profound institutional and technological changes in economic history took place in the years between 1988 and 1994 (McCann, 2008; McCann and Acs, 2011). In terms of institutional changes, the period witnessed the fall of the Berlin Wall in 1989 and the beginnings of the emergence of the transition economies in Europe; the acceleration in the market liberalization and internationalization of China, following the Tiananmen Square protests of 1989; the emergence of a new South Africa after the release of Nelson Mandela in 1990; the creation of the EU Single Market in 1992; the Second Industrial Reforms in both India and Indonesia in 1991; the establishment of the North American Free Trade Agreement (NAFTA) in 1994; the new Brazilian Constitution in 1988 and the flotation of its new currency, the Real, in 1994. These institutional changes at the national and macro-regional scale have also been accompanied since the early-1990s by a huge increase in the number of sectoral bilateral investment treaties and national double taxation treaties agreed between countries (McCann, 2008, 2009a; Barthel et al., 2010). In addition, the invention of the World Wide Web, or more precisely of the http protocol by Tim Berners-Lee, has allowed for the integration on a common platform of an enormous range of previously diverse technologies and electronic media (McCann, 2008).
There are several key features of the modern era of globalization as it relates to economic geography. First, most of the global economy is experiencing a period of slow international convergence (Spence, 2011), whereby poorer countries are catching up with richer countries, both due to their internal structural transformations and also as firms in richer countries take advantage of the new trade and investment possibilities available in the lower income countries. However, the picture is complex. The majority of Africa have been excluded from the convergence process (Venables, 2010). Second, global trade has shifted eastward due to the rise of the Asian economies, and most notably China and India. Third, the three super-regional areas of integration, namely, the EU, NAFTA, and South and East Asia, have increased their share of almost every indicator of global economic activity, implying that global activity is (or, at least, was) increasingly concentrating in these three super-regional areas of integration. Moreover, all forms of international trade and multinational activity are increasingly oriented toward cross-border engagements within the same super-region (Rugman, 2000, 2005). Multinational firms also account for a growing share of economic growth, including 31 percent of total growth in the United States since 1990 (Spence, 2011). Fourth, particular city regions, the so-called “global-cities,” appear to play a critical role in the emerging patterns of global engagement, accounting for a disproportionate share of trade-related activities. Fifth, economic growth is not uniquely related to mega-city regions. Growth can be distributed across various urban systems in different ways in different countries (OECD, 2006, 2009c, 2011a,b; Centre for International Competitiveness 2008).
The changes in city and regional performance associated with the last two decades of globalization have also altered the way we view economic geography. The evidence that the prevailing growth and development theories could no longer explain empirical growth patterns has led to a thorough rethink of how economic development takes place and how it is related to economic geography. The importance of aspects such as human capital and innovation (endogenous growth theory), agglomeration and distance (new economic geography), and institutions (institutional economics) has been brought to the fore. As importantly, globalization has also drawn attention to the often neglected role of space. Globalization has made localities and their interaction more important for economic growth and prosperity (Rodríguez-Pose, 2011). Space is becoming increasingly “slippery,” in the sense that capital, goods, people, and ideas travel more easily (Markusen, 1996; Friedman, 2005), but, at the same time, increasingly “sticky” and “thick” because capital, goods, people, and ideas, despite being constantly on the move, tend to remain stuck in large agglomerations (Markusen, 1996; McCann, 2008; Rodríguez-Pose and Crescenzi, 2008). Consequently, globalization has made space and place more rather than less important. The unique aspects of a locality and the ability to create and strengthen a comparative advantage are at the heart of economic development and success. Many scholars have put the capacity of territories to root economic activity in the local social, institutional, and economic fabric at the heart of economic success (e.g., Storper, 1997). In other words, the specific institutional arrangements and constructs of any space are increasingly regarded as the filter through which economic activity takes hold of different territories (Acemoglu and Robinson, 2000; Rodrik et al. 2004; Acemoglu and Johnson, 2006a,b; Storper 2011).
Yet, there has been one area of economic development and growth which has remained more or less still in time, impervious to the changes in the location of economic activity and in the thinking about growth and development: that of development policy (Lagendijk and Conford, 2000; Pike et al., 2006). Economic development actors in different parts of the world have frequently ended up designing and implementing isomorphic policies, that is, “formulating similar, if not identical, development strategies” (Chien, 2008, p. 274). Hence, the rise of a new territorial structure and of new theories has hardly represented a dent in the course of the development policy tanker. With few and relatively minor exceptions, traditional development policies continued to rely on 1950s growth and development theories and, across the world, remained cut from the same cloth: supply-side strategies—or, conversely, excessive demand-side strategies—based on a sectoral rather than a territorial dimension. Decision making was mainly top down, with mixed, integrated, and/or bottom-up approaches virtually ignored and with a tendency to depend on state aid, financial support, incentives, and subsidies as key elements of the strategy. Frequently, these policies relied on the imitation of successful development strategies applied in very different contexts (Storper, 1997; Pike et al. 2006 and 2007). The outcome of this was that, largely, the same solutions tended to be applied to similar problems in different places, without any real consideration of the specifics of the wider regional and local context (Chien, 2008).
Development policies have until now generally remained instruments for the provision of infrastructure—roads, railways, sanitation, water, and the like—and state aid. State-aid-based industrialization and inward investment strategies aimed at supporting and attracting large firms to territories with a weak industrial fabric have been as popular a strategy as infrastructure investment. The appeal of these two development axes has been their simplicity, tangibility, and popularity. Building roads and sanitation is not just a precondition for development, but also something demanded by society, highly visible and extremely attractive for decision-makers. Roads and other type of physical infrastructure can be built relatively quickly and allow for ribbon-cutting right before the elections. But too much emphasis on top-down, supply-side, “one-size-fits-all” quick-fixes eventually resulted in unbalanced policies, only relevant to the formal sector, and ultimately incapable of delivering sustainable development (Pike et al., 2006). An increasing body of research has tended to highlight that, even if the aggregate impact of infrastructure policies has sometimes been positive, they have often led to greater economic agglomeration, regional polarization, and to an increasing economic marginalization of many peripheral regions where significant infrastructure investments have taken place, both in Europe (Vanhoudt et al., 2000; Puga, 2002; dall’Erba and Le Gallo, 2008), as well as in the emerging world (e.g., Roberts et al., 2010 for China). Similarly, state aid and industrial intervention has wasted resources on declining industries, lame ducks, and big projects (Ulltveit-Moe, 2008). In general, these policies have struggled to cope with the more heterogeneous economic reality emerging from globalization (Roberts, 1993), often ending as “strategies of waste” (Rodríguez-Pose and Arbix, 2001).
3. THE POLICY RESPONSE: SPATIALLY-BLIND VERSUS PLACE-BASED APPROACHES
It has taken long for international organizations involved in development to react to the challenges of globalization and to the changes in growth and development theory. After years of “business as usual” while witnessing development problems fester and regional disparities increase, five major reports rethinking development strategies from scratch were published between 2009 and 2010. Such a proliferation of reports may be just a coincidence, but it is more likely to reflect a collective reaction to the growing unease with the way development policies were conducted their consequences. The World Bank and the so-called Washington consensus had become a popular target (Stiglitz, 2002; Fine, 2003; Rodrik, 2006), but the European Union's regional development policy, the largest concerted effort aimed at achieving territorial cohesion, was also not without critics (Boldrin and Canova, 2001; Rodríguez-Pose and Fratesi, 2004; dall’Erba and Le Gallo, 2008). A response was long overdue.
Perhaps, the most influential of these reports has been the World Bank's (2009) World Development Report Reshaping Economic Geography. The World Bank's (2009) report is firmly grounded in the new economic geography theory, which advocates the advantages associated with the agglomeration effects of large cities: development and growth will be unbalanced and attempts to spread economic activity will not only reduce poverty, they will also undermine growth and prosperity (World Bank, 2009). It also resorts to a Rostovian vision of economic history, which predicts that today's emerging countries will follow the past trajectories of developed states. There is, thus, a path that according to the World Bank (2009) report, all territories must tread and this path involves a greater concentration of economic activity in dynamic poles.
The report proposes a “development in 3-D” idea as its policy tool (World Bank, 2009, p. XIX). “Development in 3-D” is based on three development dimensions (density, distance, and division) and three types of solutions—the 3-i's—to overcome each dimension (institutions, infrastructure, and interventions). Each dimension has its own solution. Spatially-blind institutions should be fundamentally targeted for problems of inadequate density or agglomeration; connective infrastructure will overcome distance problems; and regional (meaning supranational) economic integration and spatially targeted intervention for three-dimensional problems and division (World Bank, 2009, pp. 23–24). In particular, spatially targeted interventions should, according to the report, be used sparingly—i.e., to address specific problems, such as the existence of large slums in urban areas—and only when “institutions for land and basic services are reasonably effective and transport infrastructure is in place” (World Bank, 2009, p. XXIV). The result is a development model based on “spatially-blind” strategies, this is “policies that are designed without explicit consideration to space” (World Bank 2009, p. 24) as the most effective way of generating efficiency, guaranteeing equal opportunities, and improving the lives of individuals wherever they live and work.
Although being primarily concerned with the reform of the EU budget, the highly influential 2004 Sapir Report An Agenda for a Growing Europe came to conclusions that were not too dissimilar from those of the World Bank (2009). This independent report recommended that in order to promote intra-EU convergence, EU Cohesion Policy should primarily target member states—and, in particular, new accession countries—rather than subnational regions. Along with national targeting of Cohesion Policy funding, the Sapir report also posited that there should be an agenda for institutional and governance reform in these new-accession countries, alongside additional sector policies focused on the development and adoption of information and communication technologies (ICTs) across the whole of Europe and other knowledge-based sectors.
The 2004 Sapir Report shares some features of the 2009 World Development Report in that the authors of both reports promote space-neutral intervention, with a primary emphasis on institutional reform. Indeed, there is no economic geography per se to speak of in the 2004 Sapir Report and EU policies are advocated primarily at the sectoral level. However, there are also fundamental differences between the two reports. First, the Sapir Report makes no comments or recommendations regarding urban growth, whereas this is a central tenet of the 2009 World Development Report. Second, Sapir et al. (2004) advocate funding to less developed parts of Europe, whereas the 2009 World Development Report would eschew such policies. Finally, the 2009 World Development Report takes no position whatsoever on the development role of ICTs, which are a central issue in the Sapir Report.
In contrast to the space neutrality of these two reports, the remaining four reports adopt a fundamentally different position: space matters and shapes the potential for development not only of territories, but, through externalities, of the individuals who live in them. Consequently, development strategies should not be space-neutral, but, as underlined in all four reports, placed-based and highly contingent on context. The place-based approach has a long intellectual history to it (Bolton, 1992). However, the application of the argument is particularly appropriate to the context where economies are experiencing major transitions toward new equilibria, as with the current era of globalization, in which the short-run and medium-term transitions processes may heavily influence the long-run outcomes (Thissen and Van Oort, 2010).
In essence, the place-based approach has two fundamental aspects to it. First, the place-based approach assumes that geographical context really matters, whereby context here is understood in terms of its social, cultural, and institutional characteristics. As such, a space-neutral sectoral approach is regarded as inappropriate; what are apparently space-neutral policies will always have explicit spatial effects, many of which will undermine the aims of the policy itself unless its spatial effects are explicitly taken into consideration. Second, the place-based approach also focuses on the issue of knowledge in policy intervention. Who knows what to do where and when? Underdevelopment traps that limit and inhibit the growth potential of regions or perpetuate social exclusion are the result of a failure of local elites to act and can only be tackled by new knowledge and ideas: the purpose of development policy is to promote them through the interaction of those local groups and the external elites involved in the policy.
Perhaps, the best known of the reports advocating place-based strategies has been the independent Barca (2009) report An Agenda for a Reformed Cohesion Policy. This fundamentally theoretical report posits that place-based policies are the best way to tackle the “persistent underutilization of potential and reducing persistent social exclusion” (Barca, 2009, p. VII) in all areas of Europe. Drawing from an institutional framework, the report traces underdevelopment in a place to the lack of either the capacity or the willingness of its local elite or to the centrifugal effect of agglomerations promoted or seconded by public interventions in other places. Exogenous policy action is seen as a way to trigger endogenous changes. A balance is then called for between exogenous and endogenous forces, by which local actors set targets and design projects, while the external “development agency” sets the general conditions that the former must follow and tailor to specific places.
Albeit starting form a radically different position, the OECD (2009a) report How regions grow reaches a remarkably similar conclusion to the Barca (2009) report. Rather than building on theoretical arguments, it uses a barrage of empirical analyses, including spatial econometrics in order to capture spillovers, to conclude that the objective of development intervention is to promote growth in all regions, as all regions display growth and development potential (OECD, 2009a). The result of this approach is a string of place-based interventions—defined as integrated regional policies (see also Pike et al., 2006)—co-ordinating infrastructure provision, with schooling, business development, and the promotion of innovation, as a means to achieve both greater local development and, through geographical spillovers, greater aggregate growth (OECD, 2009a). A further OECD (2009b) report—Regions matter—stresses the individual characteristics of regions and their place specificity and proposes to go beyond “one-size-fits-all” development approaches. Once again, it underlines that opportunities for growth exist in every region and the role of development intervention is to mobilize regional assets and exploit synergies (OECD 2009b).
A final influential report endorsing place-based approaches has been the Corporación Andina de Fomento's Desarrollo local: hacia un nuevo protagonismo de las ciudades y regiones (Local development: toward a new protagonism of cities and regions) (CAF, 2010). This report looks at development problems from a Latin American perspective and concludes, in the same vein as the Barca (2009) and the OECD (2009a,b) reports, that one of the answers to development problems is the promotion of integrated policies for each territory, with special attention to the need of cities, city regions, and regions. The CAF (2010) report emphasizes the role of local and regional governments and institutions in the development process and stresses the potential of local governance and “soft” institutions as key factors in promoting sustainable development (CAF, 2010).
The paradox emerging from this sudden burst of interest in the fundamentals of development intervention is that after decades of lack of change, development scholars, and practitioners are proposing—and are also therefore confronted with—two radically different paradigms: spatially-blind versus place-based approaches.
For those defending spatially-blind approaches, intervention regardless of context is the best way to resolve the old dilemma of whether development should be about “places” or about “people.” Advocates of spatially-blind approaches underscore that agglomeration and encouraging people mobility not only allows individuals to live where they expect to be better off, but also boosts individual incomes, productivity, knowledge, and aggregate growth (World Bank, 2009, p. 77, 135). From this perspective, spatially-blind policies are also seen as “people-based” policies, representing the best approach to deliver an improvement in people's lives and to guarantee equal access to opportunities, regardless of where they live. The reason is that it is assumed that, by encouraging mobility, spatially-blind strategies ultimately lead to a more even geographical distribution of wealth and to the convergence of lagging areas. Consequently, development intervention should be space-neutral and factors simply encouraged to move to where they are most productive—primarily in cities—as the best option to improve individual lives and foster overall economic growth (Gill, 2010).
In contrast, the place-based approach assumes that the interactions between institutions and geography are critical for development, and many of the clues for development policy lie in these interactions. In order to understand the likely impacts of a policy, the interactions between institutions and geography therefore require us to explicitly consider the specifics of the local and wider regional context. The place-based approach therefore argues that sectoral space-neutral policies, built around mobility, agglomeration, and thus, the promotion of specific urban sectors or technologies with little or no recourse to the regional context, will have important spatial implications (Hurter and Martinich, 1989). As such, the context may also end up being a backdoor but, nevertheless, key element of these policies, the implications of which must be considered right at the beginning of the policy-design process.
The emphasis of the place-based approach on the interactions between institutions and economic geography also offers some latitude in achieving development through alternative approaches to the organization of space. As such, on this argument mega-urban regions, regions with a variety of city sizes including intermediate and smaller cities, and even primarily rural regions, all have the potential to make substantial contributions to economic growth by achieving high levels of productivity in the activities in which they are best-suited (OECD, 2011a,b,c). From this perspective, the economy as a whole can reach its total output frontier by developing places of different sizes and densities, because it is the performance of the urban and regional system as a whole which is critical, rather than just the cities at the top of the urban hierarchy. Obviously, this also raises the question of whether the existence of mega cities at the top of a national urban hierarchy provides any distinct and systematic advantages for the nation as a whole over a national urban hierarchy with several intermediate or smaller sized cities at the top. As will be seen in more detail shortly, as well as empirical observations (OECD, 2006), there are also urban economic (Henderson, 2010) and institutional arguments which suggest that this is not the case.
From the place-based perspective, what this does imply is that the experiences of the now-developed countries must not be read off as the unique solution to long-term development issues, and that mega-urban regions are not the only possible growth pattern. Consequently, place-based approaches argue that spatially-blind approaches confuse correlation with causality, limiting our ability to make a leap from the observation of spatial concentration of activity to that of space-neutral policy settings in all countries.
4. CONTEXT AND PATH DEPENDENCY
In essence, the differences between the space-neutral and the place-based approaches center on the question of whether the territorial systems in evidence today are the result of a unique first-best solution to efficiency and space or rather of path dependency, sunk costs, and institutional issues. As such, the two approaches represent a different reading of the relationships between economic history and economic geography, giving rise to fundamentally different analytical approaches. A useful way of helping us to understand these differences is by reflecting on the development issues raised by two different sets of countries which are the objects of development policies, namely, developing countries and the countries of the European Union. After examining these different contexts and the issues they uncover, we will then be in a position to shine a clearer light on exactly what issues are at stake.
We begin by considering the countries which are the primary focus of the World Bank (2009) report: the developing and underdeveloped worlds. One of the most profound insights to emerge from the 2009 report is that even fundamental institutional reform in many of these countries, as has been long advocated by the World Bank itself, is no guarantee whatsoever of development (Barca and McCann, 2010) because of the complex problems associated with economic geography (Venables, 2010). If countries are too small or geographically fragmented, lack of international coordination and integration will mean that the impacts of institutional reforms will be limited. The reason is that they will be unable to sustain the agglomeration and home market effects required for growth (Collier, 2007). Second, where developing countries are indeed large but where institutions are insufficiently developed, it may well be the case that urban expansion is the only realistic policy option for overcoming institutional problems and promoting growth and development.1
Importantly, these urbanization arguments rest implicitly on the admission that urban growth is required precisely because lack of institutional quality or capacity curtails the feasibility of national institutional reform and precludes any real welfare compensation effects between income groups. Yet, if interregional mobility is primarily related to average rather than marginal wage differences, then destination cities will grow too large, origin regions will contract too rapidly, and many of the largest capital cities will suffer from severe congestion effects. These congestion problems will increase the veracity of the arguments and the decision-making influence of the capital city elites favoring local infrastructure enhancement schemes. As such, at the inter-regional level (Henderson, 2010; Kim, 2011) or at the intra-urban level (World Bank, 2003), the often overgenerous capital city infrastructure provision becomes an outcome of the political bargaining power of local elites rather than a result of market processes. These political economy processes thereby lead to capital deepening and agglomeration enhancement in the favored core areas and capital depreciation and dereliction in the less favored regions of origin, or in the poorer urban neighborhoods. Moreover, once they are set in motion, there is almost no way to reverse the asymmetric fixity effects of these inter-regional or intra-urban capital allocations.
Even before 2009, the longstanding view of the World Bank has been that context does not matter—in the sense of context being defined by place-specific social, cultural, technological, or industrial characteristics—and that well-functioning institutions and market mechanisms are all that is required for growth (World Bank, 2003). In terms of geography, the implicit corollary is that well-functioning spatial factor adjustment mechanisms are necessary in order to foster growth, irrespective of the spatial outcomes. However, to say that mega-urban growth should be encouraged acknowledges precisely the opposite, namely, that context really does matter, because if that was not the case, then institutional reforms alone would suffice for growth (Rodríguez-Pose, 2010). In developing countries where urban growth is recommended, cities may be needed not only in order to try to solve the problems associated with the generation, acquisition, and sharing of knowledge and resources, but, more importantly, because local institutions do not work properly and cannot reasonably be reformed, due to the entrenched and longstanding self-interests of the political elites.
Today's World Bank analysis of the policy implications of economic geography therefore appears to be somewhat at odds with the long-term position advocated by the same organization over several decades. Indeed, it seems strange that an organization that has focused so heavily and for so many years on institutional reform is now arguing that urban growth might be the main mechanism for development, precisely because institutional reform alone has little potential of success. Following the traditional approach and experience of the World Bank in identifying areas of institutional success and failure, it would appear that a more sensible approach would have been to advocate policies that build on this knowledge and deal with institutions in a wide variety of different types of places. Because development can be achieved in different spatial settings and arrangements, and less hierarchical urban systems may be more easily governable than ones with mega cities at the top, building on this knowledge exactly as argued by the place-based approach may have been more constructive.
Developed Countries: The Case of the European Union
The European Union provides a very different example to developing countries of how context and institutions affect both the logic of development and also the development policy. Even among developed countries, there are substantially different patterns and scales of urbanization and different urban hierarchies (OECD, 2006). Yet, while the World Bank would argue that context does not matter for development, it is clear that in the EU, there are opposing arguments that are at least as persuasive. These arguments relate to the various different institutional, legal, geographical, and political aspects of the place context, which provide different challenges and possibilities for development policy. In particular, the EU is a prime example of how an international territorial system emerges out of increasing economic integration among nation states with relatively similar levels of development but with different social, institutional, and technological features in regions. The result of this combined general and specific process is a continental urban system with a much greater role for smaller and medium-sized cities.
The debates in Europe regarding the case for regional policy, or Cohesion Policy as it is known, take place in this unique economic history-economic geography context, the logic of which is increasingly driven primarily by the EU Single Market, a market which is internal to the European Union, but which operates across 27 different countries and more than 30 different legal and institutional settings. It must also be noted that European development intervention takes place in an environment in which 17 members of the EU have engaged in an Economic and Monetary Union (EMU) and given up their respective currencies in favor of the Euro. By surrendering their monetary policy, it could be argued that they have given up one of their key instruments for development in favor of what can be termed as a spatially-blind type of intervention. Yet, each European member state, whether in EMU or not, retains its own independent fiscal policy, also a powerful development tool.
Importantly, and in contrast to the countries which were the focus of the 2009 World Development Report, is the fact that in the EU, the vast majority of these highly heterogeneous institutional environments are very good. For almost all of the rich Western European EU-15 countries which were members of the EU by 1995, the picture is broadly one of distinct but solid institutional environments, which have longstanding traditions and deeply-rooted cultural underpinnings developed over centuries. In contrast, for the EU-12 most recent accession countries, the picture is one of relatively poor institutions, many of which were destroyed or undermined under socialism, and which once again are highly heterogeneous. Institutional quality allied with institutional heterogeneity is one of the fundamental features of the EU. While World Bank recommendations traditionally argued for institutional reforms to promote development, rather than considering the role of geography, reforming poor or nonexistent institutions can often be easier than reforming strong but inappropriate institutions. This is a major aspect of the complex institutions geography context in which EU policies operate.
A second issue relating to institutions and geography in the EU has legal roots and concerns the fact that what is possible in terms of EU policy depends on the legal basis and legitimacy of the policy. Both the possibilities and limits of policy within the EU are defined by the various EU treaties, and most importantly, the Lisbon Treaty which codified and integrated all previous treaties. Moreover, this relates to both international activities within the EU and also to many intranational activities. Yet, the European Union is not a state in the sense of a country, and while all EU member states pool various aspects of their individual sovereignty, the level and nature of this pooling differs between countries. Membership of the common currency, the Euro, and the creation of a borderless internal area in the Schengen agreement are clear examples of institutional heterogeneity, and state behavior is monitored via the formal–informal Open Method of Coordination. Policies can only be formally proposed by the European Commission and the actual agreed and implemented policies are the result of negotiations between the member states and also between the member states and the Commission. As such, the principles of EU treaties both make possible, but also limit, the ability of the EU to coordinate activities. EU policy possibilities therefore depend on the legal underpinnings of the EU, and simple or direct comparisons between the EU and other areas of the world are to be treated with caution. Regional development policy in any country operates within a particular institutional and governance context, and in the multicountry EU case, the complexity and specificities of this are fundamental to the nature and the operation of the policy.
A third issue has an economic geography nature. Many highly productive cities in the EU are small to medium-sized cities whose dominant competitive advantage is that they exhibit high degrees of connectivity (Ni and Kresl, 2010; McCann and Acs, 2011) rather than urban or home market scale (Bel and Fageda, 2008). Neither rapid nor massive urban growth processes are evident in Europe. Indeed, between 2000 and 2006, EU metropolitan areas of over 250,000 inhabitants only increased their share of EU GDP by 0.6 percent. In some countries, such as the Czech Republic, Ireland, Poland, Portugal, Romania, or Spain, the expansion of particular capital city urban areas has been a major feature of economic growth. In addition, in a number of the Central and Eastern European EU-12 countries, economic growth has been allied with the types of widespread rural to urban migration processes described in the 2009 World Development Report. In contrast, in many northern and western EU countries, labor mobility has been primarily an urban to urban phenomenon and not necessarily from small to large cities. Urban systems in many parts of the EU are characterized by polycentric systems whereby labor mobility and supply chains operate by flows of goods and people between a variety of different centers, rather than with respect to a largely monocentric structure more characteristic of U.S. urban areas. The heterogeneity of EU institutional systems, and particularly the legacy of widely differing national land use and transport planning systems, has major impacts on the economic geography of the EU, which exhibits enormous variations in urban structures and labor mobility patterns. The result of this variation is that labor mobility within the EU urban hierarchy is not simply a matter of movements from smaller more specialized regions to larger and more diverse regions. The most notable case here is that of London, the major destination for U.K. labor inflows, and which is a city that in many ways is more specialized than many smaller U.K. cities (Dewhurst and McCann, 2002). Instead, EU spatial factor adjustment processes vary enormously across EU regions both in terms of their speed and patterns. The result is that the structure and the systems of economic geography are different in different parts of Europe, and this diversity means that simple off-the-shelf urban expansion prescriptions of the type offered by the World Bank are insufficient to respond to the economic, social, and environmental challenges faced by Europe.
A fourth and final issue and, in many ways, the most fundamental one, is that the economic future of the EU depends on the success of the Single Market. A market in which the majority of the benefits of the integration process accrue to particular countries, while others bear the bulk of its adverse consequences, is politically unfeasible and against the principles of social and territorial cohesion enshrined in the EU Treaty. A winner takes all outcome is not consistent with the fundamental philosophical basis of the EU.
It is this very institutional, legal, geographical, and political heterogeneity which EU Cohesion Policy must respond to, and which makes the EU case so fundamentally different from the types of argument offered by the 2009 World Development Report. If the World Bank logic were to be applied to the EU, it would essentially advocate the abandoning of middle-sized cities and most rural regions and focusing almost entirely on the mega cities. However, the EU case tells us that context really matters, and the solution to EU policy challenges is not simply to encourage the growth of the biggest cities, but to use local and regional institutions to help determine alternative pathways for development.
Institutions and History
A comparison of the issues raised for development and development policy between developing countries and the EU demonstrates that context and institutions do matter when we consider economic geography. In many ways, the oscillating World Bank position between the overwhelming importance of institutions or geography actually reflects a more general debate both within economic geography and also development studies, regarding the role of institutions versus geography in economic development (Sachs, 2003; Rodrik et al., 2004). While many people take partisan positions on this, the place-based approach contends that, in reality, it is the relationships between geography and institutions, and also how these vary across space, that really matter.
In terms of institutions, in the case of developing countries, it is easy to think of places where institutional reform has not taken place and where development has not occurred. This is the typical case usually discussed in the development literature, and now the types of places where widespread urbanization is advocated by the World Bank (2009). In contrast, a place-based logic argues that a different more nuanced approach is required even in these cases. While there can be various positive economic outcomes associated with large informal settlements (Glaeser, 2011), in many developing countries, the scale of the social and environmental problems associated with the urban slums is already far beyond the institutional capacity of the state to respond to or to rectify. Therefore, the ability of urbanization to overcome rather than to exacerbate institutional problems is not at all clear, because it depends on their interactions, which, in turn, may also depend both on the level of development and also on the existing limited institutional arrangements. São Paulo is not Lagos, and Santiago is not Calcutta, and thus a “one-size-fits-all” approach to development in these cities may be wholly inappropriate. Indeed, the place-based approach argues that in certain developing countries, the existing level of development allows for a development policy specifically designed to both utilize and enhance the institutional capacity available (CAF, 2010).
As such, the place-based approach emphasizes the interactions between geography and institutions as being critical for development in many developing countries. However, it is even possible to think of advanced countries such as New Zealand, where the institutional reforms associated with deregulation have been second-to-none, but the economic geography outcomes have become even more adverse, largely because the institutional reforms acted in a manner akin to falling trade costs within a Krugman-Venables (1995)-type framework (McCann, 2009b). This once again suggests that the interactions between geography and institutions are complex, and policy-thinking requires careful consideration of the appropriate institutional and governance responses beyond simply discussions regarding deregulation.
In terms of history, as well as oversimplifying the relationships between institutions and geography, the World Bank (2009) position also implicitly adopts a Rostovian linear development assumption in its approach. This may also be problematic (Hart, 2010; Peck and Sheppard, 2010). For example, one often-heard response to criticisms of space-neutral development solution is that what is occurring today in Manila is largely the same process that occurred in London, Paris, or New York in the nineteenth century. These arguments reflect the so-called Williamson (1965) curve, which suggests that in the early stages of growth disparities increase, subsequently leveling off. Apart from noting here that there is no consensus regarding the veracity of the Williamson argument, as a recent OECD paper explains, development processes are different at different stages of development (Garcilazo et al., 2010). Even among the rich countries, the development processes have been and are highly heterogeneous. The dynamics of knowledge generation and diffusion, innovation, and growth differ substantially between Europe and the United States (Crescenzi Rodríguez-Pose and Storper, 2007) and significant differences can also be observed both within and across countries in the developed world. Moreover, the recent pace of urban growth in developing countries—having taken place over typically three decades, rather than the 100–150 years afforded to most currently developed countries—is unparalleled in history (Henderson, 2010). This puts development processes in the emerging world in a whole new historical perspective.
Simple parallels across history also do not capture the complex nature of development. History matters for context and there are fundamental differences to the context in which economic development takes place. Prior to the First World War, the 15 largest cities in the world, the mega cities of their day, were all in the eight richest countries (McCann and Acs, 2011). These were countries which, by the standards of the time, had the most developed institutional architectures and were the dominant players in their colonial empire systems, which ensured access to precious resources from other poorer parts of the world and also provided viable destinations for migration outflows in response to domestic population growth. Today, the majority of the world's largest cities are in the developing world, in countries with limited institutional architecture to speak of, with scarce and highly regulated out-migration possibilities, and in which access to resources largely depends on global companies from richer countries and international political agreements in which richer nations tend to have the upper hand. Therefore, while urbanization and industrialization are closely related throughout history (McCann and Acs, 2011), taken together, these observations ought to cast significant doubt on the World Bank's (2009) emphasis on the role of mega-urban concentration as being a natural and optimal growth path.
The problems associated with ignoring the interactions between geography and institutions, of confusing correlation with causality, of adopting a linear and a historical view of development, and of ignoring the importance of place, signal why spatially-blind strategies may be inadequate to address what are fundamentally heterogeneous development problems across space.
Pathways to Growth and Convergence
The issues raised in this paper lead to two additional questions that are relevant for all discussions of development intervention in whatever context. First, the question of the optimal pattern of development and, second, the question of what this implies for inter-regional convergence.
On the first point, this paper argues that there are multiple possible pathways and multiple spatial arrangements for economic development. The spatial structure advocated by the World Bank (2009) with mega-urban concentration at the top of the urban hierarchy is one such pattern. However, there are strong grounds for believing that alternative structures may offer equivalent or even greater growth opportunities. One reason is that the type of system advocated by the World Bank (2009) implies significant transaction and social costs associated with migration to mega-cities, costs about which we know very little about. At the same time, a less skewed urban system may offer greater possibilities for widespread institutional engagement across regions in policy formulation and implementation, with beneficial social and distributional consequences.
On the issue of inter-regional convergence, the approach advocated by this paper takes a different position to the traditional schools of thought. For some observers, market convergence happens naturally as a spatially-blind process as long as market barriers are removed, whereas the other school of thought has traditionally been that “administered convergence” is something that can only be achieved via redistribution. The arguments in this paper, however, imply that convergence should not be a principal development policy objective. Rather, a developmentalist approach aiming at maximizing the development potential of all regions should be the focus of development policy. What this implies for convergence is unclear, but the essential point is that if convergence is to be promoted, this is to be done by development rather than by redistribution.
5. WHAT FUTURE FOR REGIONAL DEVELOPMENT INTERVENTION?
The issues and questions discussed in this paper raise important issues for the logic, design, and implementation of a place-based regional development policy. The starting point of a place-based development policy is the idea that most of the knowledge needed to fully exploit the growth potential of a place and to design tailor-made institutions and investments is not readily available—whether held by the state, large corporations, or local agents—and must be produced anew through a participatory and deliberative process involving all local and external actors. The place-based approach is therefore designed specifically to identify and build on the embedded local knowledge. At the same time, a place-based approach builds on local values and the “sense of community,” while it also requires openness to values from outside (Sen, 2009). The sense of community is a form of social capital that determines the institutional environment in which development takes place (Streeck, 1991) the capacity to generate consensus and trust, to resolve conflict, and to mobilize resources (Rodríguez-Pose and Storper, 2006), the level of provision of public goods (Tabellini, 2010), and, last but not least, the local willingness to pay for development (Bolton, 1992). This type of locally embedded social capital helps to overcome free-rider problems in the pursuit of development activities (Guiso et al., 2010). Openness to outside values prevents the sense of community from degenerating into “communitarian confinement.”
In contrast, by eliminating history, context, and path dependency, the space-blind approach lends itself to simplifications, which, in turn, often derive in simplistic and not necessarily more efficient policies. This mechanistic approach presupposes that the state has both the information and knowledge to design, implement, and monitor the most adequate development strategies through either monetary policies, fiscal policies, or institutional intervention. But a lack of institution-building capacity and a sense of community by the state (Streeck, 1991; Bolton, 1992; Rodríguez-Pose and Storper, 2006; Tabellini, 2010) may undermine the assumption that the “state knows best” and therefore also limit the economic returns of a spatial top-down development intervention.
In addition, the presence of “place-specific market imperfections or externalities” (Bolton, 1992, p. 192) is likely to seriously weaken the efficiency of spatially-blind policies and to contribute to the reproduction of “one-size-fits-all” approaches to development, which have predominated in top-down development intervention and that we have criticized above. The limitations of the state in terms of specific local knowledge and its traditional incapacity to adapt its interventions to differences in local institutions and to mobilize local actors—the “sense of place” is in itself a public good (Bolton, 1992, p. 194)—lead to development policies that end up generally cut by the same cloth, disregarding local preferences, and incapable of making the most of the variation in potential across places. As a consequence, untapped local potential often remains untapped.
The place-based approach therefore argues that there are alternative pathways to development, which require attention to detail and the institutional context. Mega-urban growth at the top of the urban hierarchy, as advocated by the World Bank (2009), is just one such development option, an option which brings its own challenges with it, and an option which so far has not been demonstrated to be an optimal solution (Henderson, 2010). The World Bank (2009) has effectively given up on institutional reform as an essential part of the development process and substituted it with mega-urban growth, thereby foregoing all of the alternative pathways. In contrast, by acknowledging the limits of the central state to design good local development policies, place-based strategies recognize the need for intervention based on partnerships between different levels of governance, both as a means of institution-building and also of identifying and building on local knowledge (Pike et al., 2007).
However, any policy, irrespective of whether it is a spatially-blind or a place-based policy, which emphasizes the primary role of building on the knowledge and capabilities of the beneficiaries of the policy, always faces the principal-agent problems of opportunism and rent-seeking on the part of the policy beneficiaries, due to the inherent information asymmetries. As such, the success of place-based policies requires that local and regional actors are galvanized to play a constructive and leading role in the policy. In order to foster the potential associated with the positive engagement of these actors in both the design and implementation of the policy (Rodrik, 2005), the policy process is therefore understood as an approach where local elites and all endogenous actors interact with the external agents involved in the policy (OECD, 2009a,b; Barca, 2009; CAF, 2010). The governance arrangements are not seen as unidirectional. On the contrary, there will be a need to stress the importance of fostering collaboration between and engagement by different levels of governance, as is appropriate for the policy being adopted. Such “multilevel governance” arrangements must be vertical, in other words, they must genuinely traverse the traditional demarcation lines between local, regional, and national government, and also horizontal, in that they must genuinely traverse the demarcation lines between cross-jurisdictional areas, and also between the public sector, the private sector, and the nonprofit sector from civil society.
What would these arguments imply for development policies? For the successful implementation of place-based development interventions in highly heterogeneous contexts, it is necessary to ensure that the incentives for behavior of all partners and collaborators are correctly aligned. This requires three key elements: first, the implementation of what are known as “conditionalities,” which are binding agreements that govern the principles which underpin the relationships between the different partners; second, a clear ex ante definition of the aims and intended outcomes in terms of well-being and socioeconomic progress of the interventions and a clear selection of the appropriate outcome indicators,2 necessary to assess whether the goals and outcomes have been reached; third, a system which promotes a space for public debate by all local actors open to dissent and alternative views, and coordination and collaboration between different levels of governance and institutions. These principles have been advocated for many years by the World Bank (2003 and 2008) and the International Monetary Fund and more recently by the European Commission (Barca, 2009), as the only means by which development policy can help foster sustainable and inclusive growth. More recently, U.S. domestic development policy-thinking is also moving in a similar direction.3
The recent outburst of interest in regional development intervention has addressed the important and urgent topic of the need to adapt development policies to changes in both socioeconomic trends across the world and economic theory. This drive has provided new and interesting insights into policy, but has failed to come out with a prevailing paradigm for development intervention. While some spatially-blind or place-neutral policies may represent the best option to promote economic growth and facilitate the catch-up of lagging areas, for others, even the best spatially-blind development strategy may be undermined by poor institutional environments, and therefore, place-based options are needed.
This debate has been centered around the fundamental question of whether efficiency should be concentrated in the core or there is potential for growth and development in every territory. The place-based arguments imply that tapping into unused potential in intermediate and lagging areas is not only detrimental for aggregate growth, but can actually enhance both growth at a local and a national level (Farole et al., 2011). Raising the potential of noncore outlying areas in a synchronized manner may enhance national growth by a significant amount (Garcilazo et al., 2010). Meanwhile, major core urban centers will continue to grow without the need of significant policy intervention, although some major long-term challenges linked to the difficulties associated with the management of the environment, congestion, and social inclusion issues will deserve greater attention.
The place-based argument suggests that development strategies should thus focus on mechanisms which build on local capabilities and promote innovative ideas through the interaction of local and general knowledge and of endogenous and exogenous actors in the design and delivery of public policies (Rodrik, 2005; OECD 2009a,b), creating multisectoral policy framework involving the provision of different bundles of public goods to different localities. Therefore, assessing development policy solely on the basis of convergence criteria alone, as was done with the EU Cohesion policy for example, makes little or no sense (Rodríguez-Pose and Fratesi, 2004) since convergence does not capture the socioeconomic objectives of the policy, which is to emphasize institutional and learning behavior.
What therefore emerges from this debate is the need to make development policies more capable to respond to today's challenges, and thus, more effective and efficient than past intervention. The place-based argument implies that this can only be achieved by trying to make growth and development intervention more “place-aware” by taking into consideration the sheer variety of factors in diverse geographical locations that may affect the potential returns of intervention. Only by making policies that are both place- and people-based will a strong case for regional development intervention be made.
Note that this argument is based on the as yet unquantified assumption that national welfare in the long run is maximized by urban expansion, even when taking into account the scale of many of today's slums and their associated social and environmental costs. Moreover, this assumption says nothing about the optimal pattern of the urban hierarchy.
The principles on which such outcome indicators are to be chosen are discussed in Fabrizio Barca and Philip McCann (2011) Methodological note:Outcome Indicators and Targets—Toward a Performance Oriented EU Cohesion Policy and examples of such indicators are contained in the two complementary notes on outcome indicators for EU2020 entitled Meeting climate change and energy objectives and Improving the conditions for innovation, research and development. See: http://ec.europa.eu/regional_policy/sources/docgener/evaluation/performance_en.htm.