Focused Series »

Indo-European Origins
Northern California
The Caucasus
Imaginary Geography
Home » Elections, Europe, Geography in the Media, Geopolitics, Physical Geography

The Political Regions of Europe and the Fallacy of Environmental Determinism

Submitted by on November 8, 2015 – 10:12 am 28 Comments |  
Europe Political Orientation MapGeoCurrents reader Rafael Ferrero-Aprato recently brought to my attention an interesting map of political divisions in Europe made by the Dutch electoral geographer Josse de Voogd and reproduced by The Economist in 2014. Josse de Voogd notes the difficulties and limitations in making a map of this sort: “Some countries [are covered] in much greater detail than others and there are lots of political parties that are difficult to place ideologically. The information comes from a wide range of resources over a long time-span.” In general terms, the map seems reasonably accurate. But at the more local scale, the situation often gets too complex to be easily captured in a map of this sort. As Rafael Ferrero-Aprato notes in regard to his own country, Portugal:

Speaking for Portugal though, the red corresponds to the strongly leftists regions of Alentejo/South Ribatejo (because of the latifundium agricultural system) and Setúbal Peninsula (an industrial region). It includes also the moderately leftist areas of the north Algarve, lower Beira Interior and Lisbon. So far, so good.

But after giving it more attention, the borders are not perfect: they include south Algarve (moderately right-wing) and the city of Porto, despite it being considered right-wing. Some leftist “enclaves” are missing too, such as the peninsula of Peniche (industrial fishing) and the city of Marinha Grande (industrial).

The Alentejo, Setúbal Peninsula, Peniche and Marinha Grande were also areas of strong influence of the Portuguese Communist Party during the 1926-74 dictatorships, the only force that remained organized in the face of strong repression by the regime. As such, these regions saw numerous revolts during that time.

Germany Electoral Maps 1The only country that seems to be misconstrued on the map—at least for recent elections—is Germany. As the set of maps from Electoral Geography 2.0 indicates, German elections have recently been structured largely Germany Electoral Maps 2around a north/south division, especially those of 1998, 2002, and 2005. The 1994 and 1987 (West Germany only) maps fit better with de Voogd’s depiction, although it does seem that he unduly minimizes the left-wing Ruhr industrial area.

European right-wing populism mapUnfortunately, the interpretation of de Voogd’s cartography by The Economist is not particularly enlightening. Much of the attention here focuses on environmental determinism, referring both to the map discussed above and to another map made by de Voogd, posted here to the left. As the noted in The Economist article:

Flat areas are more right wing The flat pains of southern Sweden, East Anglia, north-eastern France, Flanders and Padania vote for right-wing populists. Hilly regions like Cumbria, south-west France and most of the Alps tend to stick with the mainstream parties. This observation is not as facetious as it may seem. According to Garry Tregidga, an historian at Exeter University, hilly pastoral areas are generally characterised by left-leaning politics. One debatable explanation is that flat crop-growing areas benefit most from economies of scale, so fathers traditionally passed on their land to the first born, reinforcing differences in wealth and creating a more hierarchical political culture. In hilly, pastoral areas inheritances were more commonly split equally, which over the generations created a more egalitarian social structure and political tradition. Another (equally debatable) explanation is that arable farms need cheap vegetable-pickers and that the consequent foreign immigration into otherwise homogeneous rural areas stokes right-wing sentiment.

Europe physical mapThe Economist author simply gets the physical geography of Europe wrong. Upper Saxony in Germany and Provence in France are correctly depicted as right-wing populist strongholds, yet they are hardly flat areas. And as the “dominant political force” map indicates, many “flat” areas generally vote for the left. Examples here include southwestern France (Aquitaine is not “hilly,” despite what The Economist claims), the lowlands of Scotland, the Brandenburg region of Germany, the plains of Andalucía, and the lower Danube Valley. And what of upland area such as the Alps, the Carpathians, the Pindus, and the Cantabrian Mountains that are accurately depicted as more “rightist” in their voting patterns? As a comparison of de Voogd’s basic political map with a physical map of Europe shows, there is simply no pan-European correlation between topography and political viewpoints.

Like most geographers, I am often perplexed by the hold that environmental determinism retains on the public imagination. Actual evidence is rarely able to dislodge such fallacies. Evidently there is something deep comfortable about the idea that landforms and climate determine how we think.

Previous Post
Next Post

Subscribe For Updates

It would be a pleasure to have you back on GeoCurrents in the future. You can sign up for email updates or follow our RSS Feed, Facebook, or Twitter for notifications of each new post:

Commenting Guidelines: GeoCurrents is a forum for the respectful exchange of ideas, and loaded political commentary can detract from that. We ask that you as a reader keep this in mind when sharing your thoughts in the comments below.

  • NogbadTheBad

    I don’t know if I’d call it “environmental determinism” but I think the Economist’s point about London (and apparently other large cities) sharply differing politically from their hinterlands is a fair one.
    In fact in the UK the difference is more pronounced now since the May 2015 general election – in London and other English cities, the Labour Party gained or at least held ground, while pretty much everywhere else went Tory (Scotland did its own thing).
    OTOH there was no discernible “geographic” pattern to support for (“right-wing populist”) UKIP – if you look at a map of where they did best, it’s a real scatter-gun (outside of London, where they got nowhere).

    • I agree that the urban/hinterland divide is significant, and especially in the British case. This is not “environmental determinism,” however, because it is not based on landforms and climates.

      Good point about UKIP. But there is a vague pattern to the UKIP vote, and it is an interesting one: somewhat stronger support both in traditionally conservative Essex and Kent, and in some traditionally leftist industrial areas, such as West Yorkshire. Also, virtually very little support for UKIP in Scotland and the Catholic parts of Northern Ireland, in addition to London.

    • Ryan Lord

      Hey, obviously your general point about London stands, but UKIP had a reasonable showing in some limited parts of London (~20%) in such places as Bexley & Sidcup, Barking & Dagenham. I’m sure there are other spots.

  • Roman

    Your Map of Slovakia if 1st Stretched, 2nd has a wrong information when it comes to left-right determination (my field of expertise is electoral geography)

    • I am not sure what you mean. Are you referring to Josse de Voogd’s map? (None of these maps are mine.) He shows regionalist domination in the Hungarian-speaking areas of Slovakia, which is basically correct. The north-central part of Slovakia, however, is perhaps incorrectly mapped, as it has supported candidates from the Christian Democratic party in recent election.

      I would be very interested in your view of the electoral geography of Slovakia.

  • mountleek

    This map is a good starting point for other more precise works. And let’s just note that left-right division need’t be the most inportant aspect in some political landscapes. Those wild guesses by Economist suggest that the magazine is a popular source of at hand news and simple analyses, not a nourishing reading as it is often perceived.

    The future maps should really indicate time-span, which should be same for all countries. And some country shapes look really bad, especially for the Czech Republic. Oh my god, is this an automatic computer-made thing? The good thing is merging all leftist and all rightist parties together. Journalist in the Czech Republic often look only at the winning party in a region and try to evaluate the right-left feelings of voters without taking into account that e.g. a winning leftist party can be outvoted by several remaining rightist parties if we summarize them.

    For the Czech Republic the maps is generally correct. I’m not sure about the district Cheb (westernmost Bohemia), but OK. I don’t know on what elections the map is based, but as rightist strongholds, I perceive these regions: southern Bohemia, western Bohemia, central Bohemia including Praha, northeastern Bohemia, Brno, Zlín, Olomouc. For me, the first round of 2013 presidential election sums it up perfectly: (purple Schwarzenberg, red Zeman). The second round made many cities and countryside more polarized for left:

    • mountleek

      There is a Moravian regionalist party, but only with low support. 2010 lower house election:

      2012 regional election:

      • Most interesting. From the Wikipedia article on Moravané:

        “The main goal of the party is as follows: “On the principle of the right of self-determination of the Moravian nation, Moravané advocates for the independence of Moravia via restoration of the Moravian legislative parliament within the territorial scope of the Moravian ecclesiastical province.”[1][2]The party strongly disfavours state centralism, and expresses anti-Czech (Bohemian) sentiment; “Prague” (both the city and Czech government) is blamed for “suppressing Moravian culture, traditions and language” and for “pauperising of Moravia”.[3]

        According to Moravané, the future of the Europe lies in dismantling the modern European nation-states and establishing new states on historical territorial boundaries. Moreover the party supports the codification and recognition of the Moravian language, traditionally considered a dialect of Czechby linguists and the public.”

        The linguistic situation is interesting as well, as there are many dialects in Moravia. According to the Wikipedia article on Moravian dialects:

        “Southeastern Moravian dialects form a dialect continuum with the closely related Slovak language,[4] and are thus sometimes viewed as dialects of Slovak rather than Czech.”

        • mountleek

          Surely two neighboring dialects are more alike than two distant dialects. Central Moravian dialects are partially similar to Eastern Moravian dialects, Eastern Moravian dialects are in turn partially similar to Western Slovak dialects etc. I personally can’t imagine what would justify including Eastern Moravian dialects among Slovak dialects. The region where Eastern Moravian dialects are spoken has been part of the Czech political and cultural area since the beginnings.

          The claim that they belong to Slovak dialects may stem from these motivations: 1) Linguistic closeness, which is for linguist to analyze and in the end find themselves with the problem which measures to choose. 2) From the point of view of the Czech literary language we can hear some Slovak-like traits in Eastern Moravian dialects. The other way round, I can imagine that from the point of view of the Slovak literary language, which is based on Central Slovak dialects, some Czech-like traits can be heard in Western Slovak dialects. 3) Some people writing about Moravian topics on the internet are eager to present as many assertions of “Moravians’” distinctiveness from “Czechs” as possible. In fact, Moravian-related articles on Wikipedia contain a whole lot of ill-conceived, misinterpreted and far-fetched claims.

          I myself support renewing of Moravia, though not necessarily as an autonomous unit.

          • “I personally can’t imagine what would justify including Eastern Moravian
            dialects among Slovak dialects. The region where Eastern Moravian
            dialects are spoken has been part of the Czech political and cultural
            area since the beginnings.” — that may well be so, but the area where the Moravian-Slovak dialect is spoken, “Moravian Slovakia” is apparently quite distinctive (from the rest of the Czech areas) in folklore, music, wine, costumes and traditions:

            As for the dialects, there is indeed a dialectal continuum from Bohemian Czech in the west to the Eastern Slovak dialects, which in turn form a dialectal continuum with Rusyn:

            While some people might want to consider the Eastern Moravian dialect as a dialect of Slovak, most linguists treat it as a (distinctive) dialect of Czech (see e.g. Roland Sussex & Paul Cubberley, The Slavic Languages, Cambridge University Press, pp. 533-535). However, Eastern Moravian does share several features with Slovak (in opposition to both Standard Czech and Bohemian dialects), such as the use of long l and r (sorry I can’t do diacritics here), the shortening of long vowels, and others (see the reference above, p. 534).

            The entire issue is quite more political than it is linguistic, however: the dialectal continuum is pretty evident, and all in all, Czech and Slovak don’t differ as much as, say, Polish and Sorbian (Sussex and Cubberley, p. 55). Most of the differences between Eastern Moravian and other Czech (or for that matter Slovak) dialects are in the realm of phonology (and less in morphology). As far as I can tell, there aren’t significant syntactic differences, such as those found among Scandinavian dialects or Italian dialects.

          • mountleek

            The region is quite distinctive indeed.

            I don’t think that saying “A is a dialect of B” is a good choice of words. It implies that a standardised language consists of dialects. Often the standard language is based on a dialect. Is this perhaps a better wording? “A is a dialect of the B language system (or the B language group)” or “A is a B[adjective form] dialect”.

            The region of Eastern Moravian dialects (and particularly Moravian Slovakia in southeastern Moravia) is in my view special from the geo-linguistic angle. Proper Czech dialects (Bohemian dialects) are centered around the central Czech power centre that controls whole Bohemia. Central Moravian dialects too seem to be contained in a naturally power-centered area. (In this case I can think of two major natural power areas.)

            On the other hand we can perhaps think of Moravian Slovakia as a fringe region of the Czech (in the broad sense) power area. As a region whose natural area extends beyond to Slovakia, but which has been controlled by the Czech power and thus somewhat cut off from its greater hinterland. The better expression of natural geographic relationships may be the Moravian state (“Great Moravia”) of the Early Middle Ages, which was centered probably in Staré Město near Uherské Hradiště in current Moravian Slovakia. After its demise, its core areas were in the course of time incorporated into the Czech, Magyar and Austrian states.

            Coming back to the region, in Czech it’s now commonly called Slovácko (the adjective is “slovácký”), or less often Moravské Slovensko (Moravian Slovakia), which was used before. Its inhabitants are called Moravian Slovaks (sg. moravský Slovák, pl. moravští Slováci). The adjective “Moravian” is used here to specify and differentiate from Slovaks living in the Kingdom of Hungary (sg. uherský Slovák, pl. uherští Slováci; “Uhry” and “Uhersko” are Czech names for Kingdom of Hungary until 1918).

            Considering all the above, it in fact shows that the adjective “Slovak” (“slovenský” in Czech) can be comfortably applied to Slovácko. Consequently, the dialects of Slovácko may be included among Slovak dialects.

            A map of dialects in the Czech Republic:

            Apart from the Moravian Slovaks and the “uherský” Slovaks (Slovaks of Hungary), there were also Slovaks of Lower Austria. They are mentioned in ethnographic literature and visible on ethnic maps of Austria-Hungary in the northeasternmost part of lower Austria. Eventually they became Germanized. Try look at an online map here:

          • When linguists say “A is a dialect of B”, we really mean that “A is a dialect of the B language system (or the B language group)” or “A is a B[adjective form] dialect”. It’s just a shorter way to say it. It’s absolutely not saying that A is a dialect of *standard B*.

            As for determining what A is a dialect of, B or C, linguists consider linguistic features, not the geography of political power, least of all of historical geopolitical power.

  • Roy Coleman

    “Evidently there is something deep comfortable about the idea that landforms and climate determine how we think”

    As far as I’m aware, climate change was intimately related to the increase in hominid brain volume and that geographic variations in diet were crucial to this growth.

    I’m not sure what your point is.

    • If I may jump in here, the point Martin is making is that living in different physical environments and climates (in this case, mountains vs. plains) does not affect people’s political inclinations, voting patterns and such like…

      • Roy Coleman

        Perhaps. It depends entirely on the ‘geographic’ variables one chooses to examine.
        It’s hard to deny that Europe’s postwar political ‘climate’ was milder in the west than in the harsher east, or that liberal democracies have manifest in high latitude environments and have been almost exclusively absent in equatorial regions.
        Civil rights and political freedoms too correlate with a major geographic variable, the number of borders a state has,
        and impact core/peripheral developmental issues – landlocked states with more neighbours are demonstrably more troubled.
        (Mongolia vs Afghanistan for example). Ethno-linguistic diversity, which as you know has strong components of geographic distribution, also correlates with indices of war and ‘illiberal’ government.

        These are not trivial dynamics yet currently academics seem unwilling to develop a more complex nuanced view of how these interact.

        • Thank you for expanding on your point, but I can’t agree with much of what your write. The fact that “Europe’s postwar political ‘climate’ was milder in the west than in the harsher east” was a result of Stalin’s victories, not climate. Political democracies are found in a number of tropical countries, with Costa Rica forming a good example. It may be true that, as your write, “civil rights and political freedoms too correlate with a major geographic variable, the number of borders a state has,” but correlation does not equal causation, and beside that, this is a feature of the political environment, not the natural environment. The same is true in regard to ethnolinguistic diversity.

          I do agree, however, that all of these topics are non-trivial, and I welcome research that examines their interrelations. But I doubt that we will find clear cases in which the natural environment determines the ideas that people hold.

          • Roy Coleman

            I’m not arguing for any simplistic environmental ‘determinism’. Rather it is influential academics who are conflating correlation with causation in service of their various political agendas.
            Nathan Nunn’s almost impenetrable “The Blessing of Bad Geography in Africa” should, for example, be of concern to you as a Geographer.

          • Good points. Puga and Nunn’s argument is intriguing, but it may well be, as you indicate, a conflation of correlation with causation. For readers who may be interested in this, their argument runs as follows:

            “For the world as a whole, we find a negative rela- tionship between ruggedness on income. We also find that rugged terrain had an additional effect in Africa during the fifteenth to nineteenth centuries: it afforded protection to those being raided during Africa’s slave trades. By allowing areas to escape from the detrimental effects that the slave trades had on subsequent economic development, ruggedness also cre- ates long-run benefits in Africa through an indirect historic channel. We show that this differential effect of ruggedness is found in Africa only, it cannot be explained by Africa’s unique geographic environment, and it is fully accounted for by Africa’s slave trades. On the whole, the results provide one example of the importance of geography through historic channels.”


          • Roy Coleman

            It’s not intriguing, it’s bad. Nunn advances exactly what you’d maintain is impossible – the relative ‘flatness’ of African terrain ‘determined’ its economic history !
            More problematic is that he was Wrong. (His result is acutely sensitive to the influence of small numbers, which is statistically verboten)
            It is interesting to note that bizarrely too, attempts were made to model the econometric growth of places that were Uninhabited..

          • You have obviously read this paper much more carefully than I have, as I have merely skimmed the introduction and conclusion. You may well be right that the article is simply “wrong.”

            But I don’t want to come across as saying that there is no relationship between physical geography and economic development. There clearly is such a relationship, but it is complex and varies in different times and places. What I object to is the idea that physical geography determines ideas, or culture more generally.

          • Roy Coleman

            Agreed, ideas are just that and culture is so ill-defined as to be virtually meaningless.

          • Roy Coleman

            The point at which it is reasonable to assume correlation = causation is subjective. Moveover there’s a sliding scale of ‘significance’ used in differing scientific disciplines. How does one interpret this then independent of ‘ideation’ or ‘culture’?
            Sex-ratios are unnaturally skewed on the Arabian peninsula probably due to large scale femicide. This has found political expression at a global level with Daesh now seeking the destruction particularly of a Europe which ‘coincidentally’ has had ‘gay’ leaders and currently has a dozen female heads of state..

          • Causation is not a subjective decision. It’s a logically provable notion.

          • Roy Coleman

            And can be proven mathematically. Human characteristics though generally display non-linear distributions – the bell curve. If there’s a 95% correlation between lung cancer and smoking are we correct to assume causation? We can in fact do so, logically, as there are known SNPs which may account for why the remaining smokers do not. The point is that causation does not require a 100% correlation between variables, it is context dependent.

          • There are complex cases (like cancer in your example) where more than one factor can cause the same (or similar) outcome. It doesn’t change that there is no subjectivity here…

          • Roy Coleman

            There is currently no way to ‘objectively’ measure the contributions of multiple factors that lead to outcomes of human behaviour. (A driver may be the
            proximate cause of a road accident and a wet road the distal cause but how do we estimate the proportions mathematically?)

            Current thinking by Acemoglu et al is that Institutions, Culture and Geography in that order of importance contribute to geo-political outcomes. Yet as Besley has noted political and economic scientists have been
            positively hostile to the inclusion of ‘Quality of Government’ as a determinant factor, which arises from a historic ‘subjective’ political bias.

    • By “how we think” I mean the content of our ideas. That is a very different mater from brain volume,