![Sample our Urban Studies journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110826/TOC-Subject-Sample-2021-Urban-Studies.jpg"})
ABSTRACT
Written as the statistical institutes of the EU member-states discuss the development of a common technique for delineating functional economic areas (FEAs) after people's commuting patterns, the paper expresses concern over the imposition of qualifications that distort the economic reality on the ground, and over the use of asymmetric thresholds as they are bound to produce formations with different labor-market-related diffusion levels compared to the rest. To illustrate with in example, it algorithmically groups localities and delineates Greece's FEAs, entirely on the basis of people's travel-to-work flows at four different in- and out-commuting thresholds; and finds that unless neat shapes, equivalent sizes or other features are imposed by some assumption, they are not borne out by the situation on the ground. It also estimates the country's spatial fragmentation on the basis of the recovered formations; and compares the shapes of these formations to the regional/subregional division of Greece. To the extent the two grids diverge, there is room for better targeted policy interventions. On the whole, the paper advances our understanding regarding the diffusion of labour marker phenomena and interventions across the terrain in a crisis-hit country of the EU South, thus contributing to the analysis and formulation of relevant policies.
Acknowledgment
The constructive comments of an anonymous referee are greatly appreciated. The usual disclaimer applies.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 For instance, communities and clusters of communities may have tighter economic links with places in other administrative regions than they do with places (or the central place) in their own region.
2 The Nomenclature des Unités Territoriales Statistiques is the five-tier hierarchical structure used in the EU to standardise territorial units. In Greece, administrative regions correspond to NUTS level 2 sized-districts. These are divided in prefectures that correspond to NUTS level 3 sized-districts, and municipalities that correspond to upper level Local Administrative Units (LAU1s). The latter are divided further into lower level Local Administrative Units (LAU2s). Greece has 13 NUTS level 2 sized-districts, 74 NUTS level 3 sized-districts (see ), 325 LAU1s and 6,132 LAU2s.
3 It is easier and more economical to engage in policy intervention in a place from which the policy will be spill over and reach many communities via people's ordinary economic activity and interaction, than to carry out separate interventions in each of the said communities.
4 As the reader may or may not know, GDP fell by 26.5% during 2008–17 (against +17.3% in the EU), unemployment rose from 7.8 to 21.5% (compared to the EU's respective 7.0 and 7,6%), outward migration increased (estimated at the 7.6% of all employed in 2008, during 2008–16), poverty and social exclusion increased (by about 35.6% during 2008–16, against 23.5% in the EU). (The migration estimate is based on a report in the Greek daily, Kathimerini (Citation2016), and Eurostat’s 2008–16 data. The other estimates are based on Eurostat. Intra-EU migration patterns of EU citizens are not usually tracked by Eurostat or the Hellenic Statistical Authority.)
5 For instance, bodies of water that were once seen as natural barriers, when bridged or fitted with ports may become links and gates in the economic diffusion process (e.g., Merk, Manshanden, & Droes, Citation2013). As a rule, transport and communication advances tend to extend the areas of economic activity.
6 E.g. by Fox and Kumar (Citation1965). Nevertheless, Iowa's county map remains unaltered since 1857.
7 Economic activities are not confined to built-up areas, let alone that build-up areas not only vary but also may be in continued use, seasonal use or abandoned. Commuting behaviour and transportation are neither unidirectional nor uniform. Indeed, their patterns are often complex, even vary both from town to town and from hour to hour (Hincks & Wong, Citation2010; Guérois et al., Citation2014; the sources cited therein).
8 In the absence of annually solicited data at the LAU2 level in several countries, the re-estimation of FEAs every few years may be unavoidable.
9 Regarding, inter alia, (a) income, (b) sectroral or occupational or demographic features, (c) the preferred or available modes of transport.
10 Such as (a) internet dealings between places, (b) dealings between production facilities and headquarters, (c) dealings between agricultural areas and food processing industries, (d) the transportation of raw materials, industrial and consumer goods.
11 E.g., attending higher-level education institutes, using health facilities, interacting with government agencies.
12 For instance, some processes may be simple or manually interrupted (modified) at some instance or rely on some (but not all) commuting flows, while other processes may be not.
13 The supply- and demand-side self-containment ratios are equal to TAA/RA and TAA/EA, respectively, in terms of the notation employed in Section 4.
14 For instance, the 10% ratio appears in the Belgian, and the 15% ratio appears in the Estonian and OECD's FUA methods mentioned above. Among other cases, both the 15 and 20% ratios appear in a Greek pilot-study carried out for Eurostat on the basis of the 2001 data (at a more aggregated level, hence not readily comparable to the one described here), and both the 20 and 25% ratios appear in corresponding pilot-study carried out in neighbouring Cyprus (Prodromídis, Citation2009, Citation2010).
15 The correlation coefficient, which measures the strength and direction of the relationship between any two factors, in the NUTS level 2- (level 3-) sized areas, is about 63% (60%) in the case of Χ and W, 64% (46%) in the case of X and Y, 77% (64%) in the case of W καιY.
16 To remove any overlaps (so that they are not taken into account twice or thrice) in the construction of the composite measure: (a) vector W is econometrically regressed on X and an estimate of a net Ŵ is obtained, (b) vector Υ is regressed on Χ and Ŵ and an estimate of a net Ŷ is obtained. Then, by applying the formula 9 × (element value − column's minimum value) / (column's maximum value – column's minimum value) + 1, vectors X, Ŵ and Ŷ are adjusted/scaled to the 1–10 range. As a result, the product of the three columns (i.e., the measure provided in the last column of ), takes values between 1 and 1,000. The procedure is performed separately for the NUTS level 2- and level 3-sized districts.
17 On the other end, the least (internally) functionally fragmented districts are seven on the Attic peninsula, three in Central Macedonia, three in the South Aegean Islands, two in Western Thrace, one in Western Macedonia, one in the Ionian Islands, one in Crete.