Borderlines: Economic Resilience on the European Union’s Eastern Periphery

Abstract

The concept of resilience is gaining increasing traction in economic studies. In framing economic resilience, surprisingly little consideration has been given to the role played by the geographical positioning of a region, particularly those that lie on the external borders of a political and trading bloc such as the European Union. Drawing on empirical research, this chapter considers evidence for the possible adverse effects of some forms of positionality for resilience outcomes. The results of our analysis indicate that there is an external border effect that can adversely affect resilience outcomes. However, the effect is not uniform, suggesting that it is one factor amongst a number of others that can affect the economic resilience of a region. Our results confirm the importance of considering both policy dimensions and the structural characteristics of regions.

1 Introduction

The economic costs of political borders are now an accepted part of the economic literature. With the accession of new member states to the European Union (EU) in 2004 and 2007, the EU’s external border shifted eastwards, with the movement of goods and labour eased within the newly expanded EU. The long land border that was now established with Belarus, Russia, Moldova and Ukraine was also characterised by unstable relationships in political, economic and security spheres, which had manifestations in border relations. Whilst the risk of borders propagating economic shocks is reasonably well discussed, whether the presence of a hard border has an impact on the ability of a region to respond to an economic shock has been less considered to date.

In principle, the potential for an adverse effect could be realised through restricted trade opportunities or less mobile resources across borders. This might be anticipated to limit the options for an economy seeking to respond to an economic shock. However, the presence of a border might equally open up alternative opportunities that are not present to regions that rely on trade within a single market or economy.

The economic crisis of 2007–2008 heralded the most severe and protracted economic downturn in the history of the European Union. Studies have demonstrated the complex role played by socio-spatial relations in the formation of the crisis and the contagion effects that reverberated across the world (French et al. 2009; Aalbers 2009). EU regions and member states were particularly adversely affected, although the effects have varied widely (Hadjimichalis and Hudson 2014; Hendrikse and Sidaway 2014).

Economic geographers have increasingly turned to the concept of economic resilience as a means to understand differences in the observed response of regions to economic shocks and their ability to adapt to new economic circumstances. Here, we understand regional economic resilience as an evolutionary concept, namely as the capacity of a regional or local economy to withstand, recover from and regroup in the face of market, competitive and environmental shocks to its developmental growth path (Martin and Sunley 2015; Bristow and Healy 2014a).

Quantitative studies of the economic resilience of European regions to the post-2007 crisis have yielded some valuable insights into the scale and nature of spatial differences in resilience outcomes and some of the factors influencing these outcomes (e.g. Davies et al. 2010; Groot et al. 2011; Psycharis et al. 2014; Capello et al. 2015; Sensier et al. 2016). However, these have tended to treat all regions as similarly located. The positionality of a region, in terms of its location or geographical features, has not been considered. This chapter seeks to address that gap and asks whether the presence of an external border has an effect on a region’s resilience and compares this to other territorial features, such as the presence of mountains, coasts or islandstatus. We illustrate this with reference to the EU’s eastern border.

2 Border Regions on the Periphery

Promoting economic and social cohesion has been a long-standing ambition for the EU. This has underpinned the EU’s various formulations of its regional policies (now known as Cohesion Policy) since, at least, the 1980s (Bachtler and Mendez 2013; McCann 2015). These policies have had a strong geographical dimension as most of the EU’s least prosperous regions are to be found on the periphery of the EU, with their development lagging behind a more affluent core.

This core-periphery representation of the territorial economy of the EU owes much to the seminal work of Krugman (1991). Here, Krugman argued that regions could organise into a prosperous industrial core and poorer agricultural periphery through the forces of endogenous growth and the desire of firms to reap scale economies and minimise distance to market. This places border regions, located at the edge of a common economy, at a potential disadvantage in securing levels of economic growth that are comparable to the core. With the development of the EU into a single market with significant economic core-periphery disparities, the challenge for border regions has become more marked, particularly where transport connections are less strong (Schürmann and Talaat 2000).

The academic literature also identifies how a ‘border effect’ can reduce the level of trade below that which might have been expected if there was no border present (McCallum 1995). Although the scale of this effect is debated (Yi 2003), there is a wide-ranging agreement to its presence. In addition to distance to market, factors underlying the observed border effect include home market effects, whereby consumers prefer to purchase domestically produced products, the presence of barriers to trade and the lesser movement of, or access to, ideas, labour and capital between places separated by a political border. First identified in the case of the North America-Canada border, a region with, as McCallum puts it, ‘a relatively innocuous’ border (p. 622), it is suggested that in regions with more substantive borders, the effects will be more marked. The desire to reduce such border effects partly underpins the single market ethos of the EU and also influences the EU’s approach to neighbouring regions which lie outside of its territory.

The eastern periphery of the EU forms the Union’s longest land border, with eight EU member states bordering four neighbouring countries.¹ In 2004, the EU launched the European Neighbourhood Policy (ENP) as a means of promoting its relations with states that are located close to the borders of the EU (CEC 2004). Initially, EU actions in this area were, broadly, regarded as a ‘force for good’ albeit with a ‘predisposition to ethical action’ (Barbé and Johansson-Nogués 2008, p. 81). However, more recently, observers argue that the neighbouring countries have become increasingly unstable, and economic transition has slowed down, leading to suggestions that it might be time to reset the relationship (Lehne 2014).

Whilst the eastern borders of the EU constitute just a small part of the ENP, they do highlight the range of tensions and economic shocks that can beset border economies. This includes susceptibility to particular shocks such as the closure of markets through economic sanctions, political turmoil in neighbouring countries, first destination of mass migration and the closure of borders. Each of these can have ramifications that are often accentuated in border regions that tend to be more dependent on cross-border trade than their counterparts elsewhere in a country or the EU as a whole. This highlights the interest in promoting more resilient economies in border regions. In an unusual development, we also see some cases of very particular shocks that border regions can be subject to (though not uniquely so). In 2007, cyberattacks swamped websites of Estonian organisations, including banks, ministries, media publishers and broadcasters, and the Estonian parliament. Whilst the source of the attacks has never been proven, it has been ascribed to political tensions between Estonia and neighbouring Russia. This development of modes of, what are commonly termed, hybrid warfare has led North Atlantic Treaty Organization (NATO) commentators to now describe resilience as ‘a core element of collective defence’ (Shea 2016).

3 Economic Shocks and Resilience

Resilience is a complex and multidimensional concept. Typically, the resilience of a system is described in terms of its ability to resist shocks or the speed by which it is able to return or ‘bounceback’ to a pre-shock state or equilibrium (Pendall et al. 2010). This has been used by economic geographers to assert that a resilient region is one that demonstrates the capacity to resist a shock in the first place or to recover quickly from its disruptive effects (Martin 2012). Evolutionary economic geographers extend the resilience concept further by arguing that regional resilience should also be conceived as the ability of a regional economy to adapt and re-orient or renew itself over time (Martin 2012; Bristow and Healy 2014a). As regional economies are constantly evolving adaptive systems, this ability to transform themselves is an essential feature to either avoid becoming locked into a suboptimal development path or to transition to a ‘better’ one (Hill et al. 2008; Martin 2012; Isaksen and Trippl 2014; Bristow and Healy 2014b).

Typically, existing assessments of regional economic resilience to the post-2007 economic crisis across Europe have focused primarily on the ability of regional economies to withstand the crisis in the first place or suffer limited short-term disruption to their overall economic performance. This is partly a reflection of the limited time that has passed since the crisis for data to be available against which to make such an assessment. Davies (2011), for example, examines how the resilience of regional economies varied across European countries in the immediate aftermath of the crisis, in the years 2009 and 2010. Here, the analysis of resilience is measured in terms of percentage changes in the regional unemployment rate for a relatively small cross-section of ten European countries (Austria, France, Germany, Hungary, Italy, Poland, Romania, Spain, Sweden and the UK). Groot et al. (2011) similarly examine cross-country and cross-regional variations in the short-term impact of the crisis but this time using gross domestic product (GDP) growth data for 2009 for nine EU countries. As their data are limited to national data sets, they have only a limited assessment of the regional dimension to the crisis.

In perhaps the most comprehensive assessment of the resilience of European regions to the crisis, Sensier et al. (2016) construct a method for assessing regional economic resilience based on the date at which each region individually experienced the onset of the crisis and the extent to which it had recovered by the end of 2011. Using data for total employment and for GDP, they apply this to each region in the EU, alongside Iceland, Liechtenstein, Norway and Switzerland. Their employment analysis identifies that more than a tenth (12%) of regions weathered the crisis and did not experience any fall in numbers employed, whilst almost a quarter (23%) experienced a fall in employment but, by 2011, had recovered to their pre-crisis peak. In contrast, two-thirds of regions were still to recover by 2011, divided evenly between those that had passed the trough of their downturn and those that were still to register the end of their decline in employment. An analysis of resilience based upon GDP data gave similar results (Sensier et al. 2016).

Assessments of the factors that shape regional economic resilience tend to focus on those innate characteristics that underpin the structural features of regional economies (Rose 2004). Inter alia, these include the strengths and weaknesses of regions as they enter a crisis (Davies et al. 2010). They report that the size of the available market and access to a larger external market, as well as endowments in natural resources and in physical and human capital, all play an important role in shaping variable impacts of shocks across regions. The important role of a qualified labour pool is of significance here (Bristow et al. 2014). Other factors that can be influential include the sectoral structure of regions, with strong correlations reported between economic diversity and greater levels of resilience (Davies et al. 2010; Bristow et al. 2014). This provides further support for theories that highlight evolutionary conceptions of resilience and the value of ‘species diversity’ for regional economies (Bristow 2010).

Much of the evidence available points to the important role that the industrial legacy of a region can play in shaping its resilience to economic shocks. This highlights the path-dependent nature of regional resilience and the scope for re-orientating skills, resources and technologies inherited from that legacy (Boschma and Martin 2010). In a similar vein, regions with higher levels of innovation activity appear to be able to respond to economic shocks more positively than those where innovation capabilities are lower (Bristow et al. 2014). Whether this is due to the value of the innovation activity itself or owes more to the propensity to innovate, acting as a signifier of ability to respond and adapt to changing circumstances remains uncertain.

Increasingly, academic research is also beginning to recognise the important role that governance and agency can play in shaping the resilience of regions and regional economic development (see Bristow and Healy 2014a). This relates to the quality of governance (Charron et al. 2014) and also the importance of recognising the territorial context for policymaking (Barca 2009; Bristow and Healy 2014b). However, resilience studies have, to date, paid less attention to the question as to whether territorial characteristics might, in themselves, impact upon the resilience of place.

4 Assessing the Economic Resilience of Border Regions

To assess whether the presence of an external border has an influence on the resilience of a region to an economic shock requires us, first, to analyse which regions were resilient to a given shock and which were not. Using the economic crisis of 2007–2008 as a test case is a valuable opportunity, as it is one of the few shocks that can be said to have been experienced by all regions of the EU. For the purposes of this study, we adopt a business cycle approach whereby we construct the individual business cycle for each of the 31 national economies in the study and the 281 NUTS 2 regions and 1322 NUTS 3 territories within this geography (for details of the methodology, see Sensier et al. 2016). This approach has the distinct advantage of accommodating the different time frames for when different regions were affected by the economic shock rather than assuming that all regions are affected simultaneously. The use of NUTS 3 data allows a smaller scale of analysis than is typically adopted in studies of economic resilience (which tend to focus on NUTS 2). This finer-grained analysis enables the effect of territorial characteristics to be more readily identifiable than might be the case using NUTS 2 data. Owing to data limitations, analysis at the NUTS 3 level is restricted to 28 countries (Norway and all EU member states with the exception of Croatia).

Taking the definition of resilience as the ability of an economy to maintain existing levels of economic activity in the face of an economic shock, or to recover to the pre-shock peak within a given time period, we then identify the point at which each region enters into economic downturn and the point at which it recovers to its pre-crisis peak of economic activity. We have chosen to use the level of employment in a region as a more meaningful measure of resilience than other alternatives, particularly GDP. This is partly because it is less prone to revision than GDP data, partly due to its greater robustness at the NUTS 3 level, and also because of the social value associated with being in work. There is a tendency in the minds of the public and politicians regarding the possession of a job as a strong indication of the well-being of an economy.

We measure the absolute resilience of the economy to an economic shock, rather than its resilience relative to other economies (Martin 2012), as we are interested in the extent to which territorial characteristics influence the resilience of a region rather than the relative resilience of a border economy to all others. We have also followed a convention that to be considered resilient, an economy should have recovered to its peak employment levels within three years of experiencing an economic downturn (Sensier et al. 2016). On the basis of their experience since the crisis, each economy is judged to be in one of four states, which are mutually exclusive: Resistant, Recovered, Not recovered: Upturn and Not Recovered: No Upturn.

Resilient regions (Table 12.1) are those that did not experience a downturn in economic activity, following the economic crisis (Resistant) or those that experienced a downturn in economic activity but recovered to pre-shock peak levels by 2011 (Recovered). Regions that were not resilient to the crisis are those that have not recovered to pre-shock peak levels by 2011. This category is subdivided into two further categories: those that have registered an upturn in activity levels but had not recovered to their pre-shock peak by 2011 (Not Recovered: Upturn) and those that were still to record an upturn in activity by 2011 (Not Recovered: No Upturn).

Table 12.1 Regional resilience categories

Each NUTS 3 region was then assigned to one of five territorial types, which are not mutually exclusive: a mountain region, a coastal region, an island region, a region with an external border or a region with none of these features. The classification was based on that produced by the European spatial research programme (ESPON 2014).

For each region, we examine the extent to which it is more or less likely to be resilient than would be anticipated, given the average propensity to resilience to the economic crisis. Where a territorial type is associated more strongly with a particular resilience state than might be expected, given the average distribution of resilience then a value greater than 1.00 will be recorded. The higher the value, the greater the extent to which that resilience state is over-represented. In contrast, values of less than 1.00 signal where a territorial type is less associated with a particular resilience trajectory than would be expected, given the overall distribution. Values close to or equal to 1.00 suggest that a particular trajectory is neither more nor less likely to have influenced the distribution of resilience states.

5 The Economic Resilience of Border Regions

In their analysis of the distribution of regional economic resilience, Bristow et al. (2014) identify a strong geography of resilience, clearly influenced by national patterns (Fig. 12.1). However, important pockets of recovery and non-recovery are also apparent within this overall geography. This is particularly apparent on the eastern border of the EU, where Polish regions proved to be able to resist the crisis. Elsewhere along the border, regions proved not to be resilient, either continuing to experience a decline in economic activity or having begun the path to recovery, but still to achieve pre-crisis employment levels by 2011.

Fig. 12.1

Distribution of regional economic resilience (NUTS 2, employment, peak year to 2011) (Source: Bristow et al. 2014. ESPON Database, ESPON ECR2 Project, Cardiff University, UK. Origin of data: Experian, Cambridge Econometrics, Eurostat. EuroGeographics Association for administrative boundaries)

To what extent though, does the presence of a border influence the observed resilience of regions to the crisis? In Table 12.2, we estimate the extent to which regions with different border characteristics were more or less likely to be found in one of the four resilience categories. Non-border regions have the strongest propensity for resilience and are more likely to have resisted the effects of the economic crisis. Territories with internal borders exhibit a stronger propensity to have recovered from the effects of the crisis. Those territories that have external borders exhibit the weakest levels of resilience. They have a much stronger representation amongst regions that have not recovered than might otherwise be expected, strongly suggesting the presence of a ‘border effect’.

Table 12.2 Border regions and regional resilience

A similar analysis for regions with mountainous and coastal characteristics also demonstrates that such regions tend to have proven less resilient to the economic crisis (Table 12.3). Mountainous regions form a higher proportion of regions that have not yet recovered from the economic crisis and areas with low, medium, high or very high coastal populations also make up a disproportionate share of regions that had still not recovered their peak employment levels in 2011. In contrast, regions that resisted the crisis, or have since recovered, are more likely to be non-mountainous or to be found in non-coastal areas.

Table 12.3 Regional resilience and mountain and coastal characteristics

However, this simple correlation of territorial type with resilience outcomes does not take into account the possibility that territories with particular characteristics appear to have weaker levels of resilience outcomes simply because they are disproportionately located in countries where overall levels of resilience are already weaker. When the national context is controlled for, a more complex picture emerges. Overall, it appears that in around a third of countries (10/28), the territorial characteristics of regions may have some influence on the observed level of resilience. However, there is no consistent pattern to this, as in each case there are examples of where the same characteristics are associated with different resilience outcomes. For example, in some countries, border regions have proved more resilient than the national average, whilst, in others, they have proved less resilient. We illustrate this with two examples from countries on the EU’s eastern border: Finland and Poland (Table 12.4).

Table 12.4 Varying circumstance by country

In the case of Finland, of the 20 NUTS 3 regions, 11 are classified as coastal (coast), 1 as island (island), 7 as external borders (ext. border) and none as mountainous (mtn). In Poland, of the 66 NUTS 3 regions, 4 are classified as mountainous, 8 as coastal, none as island and 15 as external borders. In Finland, no region with an external border was resilient to the crisis, whereas in Poland almost half of the external border regions resisted the crisis, and 9 of the 15 proved to be resilient.

One interpretation of the data could also be that territories of a particular type appear to have weaker levels of resilience outcomes simply because they are disproportionately located in member states where overall levels of resilience are already weaker. In order to control for this, we consider the extent to which regions under- (or over) perform against the measured resilience of the national economy.

From our initial analysis in Table 12.4, it is a simple step to determine the proportion of regions in each Resilience category. This provides a means to assess whether regions of particular territorial types are over or under-represented in each Resilience category, and so to identify where territorial characteristics might be associated with higher or lower levels of resilience outcomes than could be expected. This is illustrated for the cases of Finland and Poland in Table 12.5.

Table 12.5 Assessing the relative situation of regions

Here, we see that in the case of Finland, regions with external borders are over-represented as a proportion amongst those regions that were still experiencing a decline in levels of employment in 2011. In contrast, the opposite was true in Poland, where regions with external borders were slightly more likely to have resisted the crisis than the average. Undertaking this analysis for each of the countries on the EU’s eastern border (including Norway) (Table 12.6) suggests that in four of the nine countries, regions with external borders proved to be less resilient to the crisis than might have been otherwise expected (particularly so in Finland and Hungary). In Poland, the presence of an external border appears to be positively related to a stronger level of resilience to the crisis, and in Romania, the picture is mixed. In three countries, the presence of an external border does not appear to have a significant influence on the propensity of a region to be resilient.

Table 12.6 Comparing the resilience of border regions within countries

In an extension of this analysis, we can also ask whether the presence of challenging geographic features, such as an external border, might affect the overall resilience of the national economy itself. This might account for the weaker resilience outcomes identified earlier, and it may help to explain why regions with external borders (or other challenging territorial characteristics) are more likely to be found in countries with lower resilience outcomes. In Table 12.7, we summarise the results of a similar analysis to that undertaken earlier. Overall, a more favourable territorial composition does appear to be important for states that resisted the crisis, with low proportions of regions in all territorial categories, although this finding is based only on three states. Similarly, the presence of more regions with external borders appears to affect the overall resilience of the member state concerned. However, a higher prevalence of mountain, coastal or island regions does not appear to be a significant factor affecting the ability of states to recover, following the onset of an economic downturn.

Table 12.7 The effect of territorial characteristics on national resilience

6 Conclusions

The reverberations of the post-2007 economic crisis have had significant impacts across the EU. The fact that all EU economies were exposed to this same shock has provided an opportunity to develop our understanding of various influences on the economic resilience of regions. In this chapter, we have considered the role that territorial characteristics play in shaping the observed resilience of regions. We have focused on the role of external borders, with particular reference to the eastern periphery of the EU. To set these results in context, we have considered the role of other territorial characteristics, such as mountains, islands and coastal areas.

Our results demonstrate that regions with external borders tended to be less resilient to the economic crisis than were regions with no national borders, or where these borders were internal to the EU. Mountainous regions also tended to be less resilient to the crisis than non-mountainous regions. The picture for coastal regions is more mixed. However, further analysis suggests that this effect is a feature of these regions being more likely to be located in member states where overall levels of resilience were low. Thus, it is likely that at least part of the effect is the result of the national economic context rather than the nature of the territory itself.

In the case of the eastern periphery of the EU, we find that in just under half of the countries concerned, the presence of an external border is associated with worse resilience outcomes during the crisis, and in one case, the presence of an external border is associated with improved outcomes. In the other four cases, the presence of an external border did not appear to affect the outcomes experienced or the effect was mixed. A critical finding of the study is that the presence of an external border can have an adverse impact on the resilience of the national economy. Thus, the effect is felt not just at the regional level but also at the national level. There was no comparable national effect for the presence of mountain, coastal or island regions.

The results of our analysis indicate that there is an external border effect that can adversely affect resilience outcomes. However, the effect is not uniform, suggesting that it is one factor amongst a number of others that can affect the resilience of a region. Wider policies are also likely to affect the magnitude of any border effect and could act to create a positive effect, as well as mitigating negative effects. This serves to emphasise the importance of considering the policy dimension in resilience studies, as much as the structural characteristics or of regions.

Our work also indicates that studies of resilience would benefit from a stronger reflection of geopolitical considerations. The rising significance attached by bodies such as NATO to the theme of resilience highlights the relevance of this matter. It is also a theme that appears to be growing in significance as economic and political tensions in neighbouring regions raise questions as to the future of the European Neighbourhood Policy and the Eastern Partnership. However, our knowledge of how geopolitical attributes affect the resilience of regions is at a very early stage, suggesting room for further research, which takes into account the particularities of individual places as well as the broader context.

Finally, and perhaps most telling, is the complex and recursive relationship suggested between observed regional resilience outcomes and resilience at the national level. It appears that the presence of external borders is associated with non-resilient outcomes in the national economy, which, in turn, then reduce the likelihood of border regions proving resilient to an economic shock. Understanding the complexities of this relationship will provide a fruitful avenue for further research.