http://orcid.org/0000-0002-3261-8038
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
Geography, cultural remoteness and the second nature of within-country economic development: do island regions lag behind? Regional Studies. Using a unique dataset that encompasses island regions worldwide (76 regions) and all the other regions of the countries (22 countries) to which they belong (474 regions in total), we document that the within-country gross domestic product (GDP) per capita delay of island regions(on average –21.7% of their national mean), while not explained by regional (neither geographical nor cultural) distance to the (neither global nor local) frontier of economic development, i.e., remoteness, is associated with ‘second-nature’ regional factors (human capital and institutional quality) not specific to island regions. However, the condition of being an island is strongly associated with worst second-nature characteristics.
摘要
地理，文化偏远和国内经济发展的第二天性：岛屿型区域是否发展落后？Regional Studies. 我们运用包含全世界岛屿型区域（七十六个区域）以及这些区域所属的国家（二十二个国家）的其他所有区域（总共四百七十四个区域）之独特数据集，记录岛屿型区域在国内的人均国内生产总值（GDP）之迟滞现象（一般较全国平均少了百分之二十一点七）—尽管未由如偏远性等相对于经济发展（非全球或在地）前沿区域（非地理或文化） 之距离进行解释—关乎并非岛屿型区域独有的区域“第二天性”（人力资本和制度质量）。但作为岛屿的境况，却强烈关乎最不利的第二天性之特徵。
RÉSUMÉ
Géographie, isolement culturel et seconde nature du développement économique interne: les régions insulaires sont-elles à la traîne? Regional Studies. En faisant usage d’un ensemble de données unique englobant des régions insulaires du monde entier (76 régions), et toutes les autres régions des pays (22 pays) auxquelles elles appartiennent (474 régions en tout). Nous documentons le fait que, bien qu’il ne soit pas expliqué par la distance (ni géographique ni insulaire) sur la frontière (globale ou locale) du développement économique, à savoir éloignement, le PIB interne par habitant est inférieur, en moyenne de 21,7%, est associé à des facteurs régionaux de «seconde nature» (capital humain et qualité institutionnelle) non spécifique à des régions insulaires. Toutefois le fait d’être une île est fortement associé avec les pires caractéristiques de «seconde nature».
ZUSAMMENFASSUNG
Geografie, kulturelle Abgelegenheit und die zweite Natur der Wirtschaftsentwicklung innerhalb des Landes: Hinken Inselregionen hinterher? Regional Studies. Mithilfe eines eindeutigen Datensatzes von (76) weltweiten Inselregionen und allen anderen Regionen der (22) Länder, zu denen sie gehören (insgesamt 474 Regionen) dokumentieren wir, dass die Verzögerung von Inselregionen beim Pro-Kopf-Bruttoinlandsprodukt (BIP) innerhalb des Landes (durchschnittlich -21,7% des nationalen Mittelwerts) sich nicht durch die regionale (weder geografische noch kulturelle) Entfernung zur (weder globalen noch lokalen) Grenze der Wirtschaftsentwicklung, d. h. Abgelegenheit, erklären lässt, sondern mit regionalen Faktoren der ‘zweiten Natur’ (Humankapital und institutioneller Qualität) zusammenhängt, die nicht für Inselregionen spezifisch sind. Allerdings ist die Bedingung einer Inselexistenz stark mit den schlechtesten Merkmalen der zweiten Natur verknüpft.
RESUMEN
Geografía, distancia cultural y la segunda naturaleza del desarrollo económico dentro del país: ¿están rezagadas las regiones insulares? Regional Studies. A partir de un grupo de datos exclusivo que abarca (76) regiones insulares en todo el mundo y todas las regiones restantes de los (22) países a los que pertenecen (474 regiones en total), documentamos que el retraso de las regiones insulares en el producto interno bruto (PIB) per cápita en cada país (de promedio –21,7% de la media nacional) no se explica por la distancia regional (ni geográfica ni cultural) a la frontera (ni global ni local) del desarrollo económico, es decir, el aislamiento, sino que está asociado a factores regionales de ‘segunda naturaleza’ (capital humano y calidad institucional) no específicos para las regiones insulares. Sin embargo, la condición de ser una isla está estrechamente relacionada con las peores características de segunda naturaleza.
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
SUPPLEMENTAL DATA
Supplemental data for this article can be accessed at http://dx.doi.org.10.1080/00343404.2017.1284310
ORCiD
Massimo Del Gatto http://orcid.org/0000-0002-3261-8038
Notes
1. Distinctive policies for islands were first mentioned at the time of the Maastricht Treaty (1992) in which the necessity to link islands, landlocked and peripheral regions with the central regions of the European Community (cf. Art. 154) was highlighted. The treaty also established a category of regions (‘ultra-peripheral’ or ‘outermost’ regions) that needed special treatment by the EU legislator: the French overseas departments, the Azores, Madeira and the Canary Islands. Later, the treaties of Amsterdam (1997) and Lisbon (2007) recognized the specific economic and social situation of the overseas departments, ‘which is compounded by their remoteness, insularity, small size, difficult topography and climate, and economic dependence on a few products, the permanence and combination of which severely restrain their development’ (cf. Art. 299(2)). The economic development of these areas is explicitly recognized in the policies of economic and social cohesion (see, for example, the analysis of the EU policies concerning the economic sustainability of islands by Moncada et al., Citation2010).
2. Regions are identified at the second administrative level. We include 31 ‘overseas islands’, i.e., special administrative entities, compared with the ordinary regions of the countries to which they belong (cf. the categories ‘overseas countries and territories’ and ‘most remote regions’ in official EU documents; Moncada et al., Citation2010).
3. In disentangling between first- and second-nature factors, we are mainly concerned with the extent to which they are the result of human activity. In the case of remoteness, man is responsible for technological progress and for the location of the development frontier, but not for the distance from it. Thus, we are just a step away from the idea of first nature in a strict sense. To be more rigorous, one might think of remoteness as a second-nature characteristic and talk about ‘third nature’ to refer to factors such as human capital and institutions. However, this would unnecessarily complicate the exposition.
4. The economic aspects of insularity have been mainly studied with a focus on small islands and/or microstates (Armstrong & Read, Citation1995; Armstrong, De Kervenoael, Li, & Read, Citation1998; Bertram, Citation2004; Bertram & Karagedikli, Citation2004; Feyrer & Sacerdote, Citation2009; McElroy & Kimberly, Citation2012), and from the point of view of single countries (e.g., Briguglio, Citation1995; Dimou, Citation2006).
5. These populations are capable of encoding the innovations and adapting them to a new context more easily. The reason for this is that the intergenerational transmission (vertical transmission) of values typical of a given population is a complex phenomenon and occurs first in the familial sphere. As a consequence of this, the intensity of the dissemination of values between populations (horizontal transmission) decreases with ‘cultural’ distance.
6. The phylogenetic tree is a diagram that shows the relationship between groups of progeny derived from a single parent. The term ‘branches’ describes the points where populations divide.
7. Spolaore and Wacziarg (Citation2009) formalize this step by assuming that populations inherit characteristics from their ancestor populations according to a random walk. This entails that the distance in vertical characteristics between two populations is, on average, increasing in their genetic distance.
8. This is consistent with the Fifth Report on Economic, Social and Territorial Cohesion (European Commission, Citation2010), which defines island regions as ‘one or more regions that consist of one or more islands’.
9. In detail, Shikoku, Kyushu, Hokkaido and Ryukyu are Japan’s minor islands. The minor islands of Indonesia comprise Maluku, East Timor, Bali, East Nusa Tenggara, and West Nusa Tenggara; while those of the Philippines are Visayas and Sulu.
10. For homogeneity with other measurements, the variable was subsequently rescaled in order to fall between 0 and 1.
12. Two comments are in order. First, new technologies can arguably come from a number of frontiers, and not necessarily from the one within the national borders. The choice of defining a single-frontier region per country is an approximation of such a continuous process of technological exchange. Second, the identification of the frontier regions might sound strange in some cases. For example, the Italian frontier is not Rome (located in the centre) but Turin, located in northern Italy (in the Piemonte region). Several southern Italian regions (such as Sicily) were not part of Italy in 1850, so that their distance to Turin might not seem fully meaningful. However, it is worth noting that what matters for our purposes is the subsequent evolution, given the situation in 1850. For example, Sicily was merged with the Kingdom of Sardinia in 1860, which became the Kingdom of Italy in 1861, and the capital was Turin. Although Rome became the capital in 1871, the economic core of Italy never moved away from northern Italy, which is where innovation still comes through. By contrast, some countries in the developing world became independent after 1850, and although some of them are still economically attached to their colonizing countries, their own centres/capitals suddenly became important nodes for the process of technology diffusion. This is the case of Jakarta, Indonesia, where the Dutch tactics to keep control of the region that were imposed during the 19th and early 20th centuries required that virtually all exports from all over the region were shipped through Jakarta, giving this town the overwhelming political and economic dominance that it still maintains.
15. A replication package (data plus STATA codes) for the linguistic index used in this paper, together with the full set of computed bilateral distance measures (geographical, linguistic and ethnic distances), can be downloaded from the first author's website (see http://docenti.unich.it/delgatto/delgatto_web/research.htm).
16. The most recent information available from 1990 to 2006 is used and, to make levels of educational attainment comparable across countries, the educational statistics are translated into the International Standard Classification of Education (ISCED) standard. United Nations Educational, Scientific and Cultural Organization (UNESCO) data are also used.
17. Each survey provides separate measurements for different sub-national regions within each country. This information is aggregated by first averaging the (standardized) data for all questions that fall within each of the four categories within each survey, and then aggregating the institutional variables across the different surveys to obtain one measure of each of the four categories of institutions.
19. The island effect in disappears when Galapagos Islands are included, while the significativeness of the Island dummy in is only slightly affected.
20. The number of observations in drops from 468 to 446 because one region for each country (i.e., the frontier region) is lost under this specification.
21. These regressions may suffer from endogeneity if one accepts the idea that, in the long run, geography itself, including being an island, may have played a role in determining the degree of cultural remoteness. To investigate this aspect, we run additional regressions in which we use the two measures of cultural remoteness (LingRemote and EthnRemote) as dependent variables and geographical remoteness, together with the same controls for insularity used above, as explanatory variables. While geographical remoteness is significant, entailing that including geographical and cultural remoteness in the same regression would be problematic, being an island does not seem to impact significantly the degree of cultural remoteness of the regions. This result is fully confirmed in regressions in which the additional first-nature factors mentioned above are included one by one (see Appendix B in the supplemental data online): the island dummy is never significant, while geographical remoteness always significantly affects linguistic remoteness and, in most regressions, also ethnic remoteness.
22. Consistent with Mitton (Citation2016), Institutions and Autonomy are not significant when the interacted term is not included (see Appendix G – part E in the supplemental data online). Regressions run with only human capital or only the institutional variables provide exactly the same information.
