Navigation – Plan du site

Accueilnumeros2025/2Are sparsely populated areas more...

Are sparsely populated areas more disconnected from local labor markets? Examining local migration intensities in northern Sweden

Les zones faiblement peuplées sont-elles davantage déconnectées des marchés du travail locaux ? Analyse des intensités migratoires locales dans le nord de la Suède
Dean Bradley Carson et Doris Anna Carson

Résumés

Les zones faiblement peuplées (SPA) ont récemment été décrites comme une catégorie géographique particulière avec un éventail de caractéristiques uniques. Une de ces caractéristiques est l’idée qu’elles sont déconnectées en interne, et deviennent de plus en plus déconnectées au fil du temps, étant donné que les villages sont séparés par la distance, par différents modes de développement et par des caractéristiques socio-économiques différentes. Les zones faiblement peuplées sont dépourvues des types de structures centre-périphérie, stables et réciproques, qui sont courantes dans les zones rurales plus densément peuplées, ce qui signifie qu’elles ne partagent pas les populations et les marchés du travail à l’intérieur de leurs régions et qu’elles comptent sur des sources de capital humain externes et plus imprévisibles. Cette étude analyse dans quelle mesure l’hypothèse de déconnexion s’applique dans le contexte de municipalités faiblement peuplées dans le nord de la Suède. En particulier, elle compare les intensités migratoires locales (à l’intérieur du comté) des personnes en âge de travailler dans les zones faiblement peuplées du Nord avec celles dans d’autres municipalités suédoises sur une période de 21 ans, prenant en considération les différences entre les groupes d’âge. Elle analyse également les relations entre le changement de l’intensité migratoire, la position nette de la migration, et la croissance démographique au fil du temps. Les résultats confirment que les zones faiblement peuplées du Nord sont davantage déconnectées que le reste de la Suède, particulièrement lorsqu’on observe l’intensité de la migration entrante. Cependant, elles ne sont pas constamment plus déconnectées que d’autres municipalités rurales dans le Sud, ce qui suscite des questions sur la validité de l’hypothèse de déconnexion comme un trait caractéristique distinct des zones faiblement peuplées. Nous constatons aussi qu’une connectivité accrue (par des intensités migratoires locales plus élevées) mène à des taux plus faibles de migration entrante pour la plupart des groupes d’âge et une position moins bonne en migration nette, ce qui suggère que le recrutement de main-d’œuvre externe, qui exige plus de ressources, pourrait avoir des résultats plus souhaitables que le développement de la main d’œuvre locale.

Haut de page

Texte intégral

Introduction

1The purpose of this paper is to examine the extent to which Sweden’s northern sparsely populated municipalities exist as ‘worlds apart’ from the rest of the country and from rural municipalities in the south when it comes to local migration exchanges of working age people within their counties. Previous research has identified that sparsely populated areas (SPAs) typically have small ‘local’ labor markets and struggle to access and mobilize those markets to meet labor needs (Stein, 2019; Lundmark, 2006; Strasser, 2016). According to the ‘Beyond Periphery’ model (Carson, 2011; Carson & Carson, 2014), a major challenge for SPAs is that they are internally ‘disconnected’ from those local labor markets, meaning that individual settlements within a region are so isolated from one another that they do not share common labor, migration or mobility flows. Disconnectedness emerges due to large distances, socio-economic diversity, and the relative absence of historically grown functional ties (e.g. transport, economic, social, cultural) and shared identities between individual settlements. This has implications for recruiting and retaining key workers (and resident populations generally) and consequently the costs of economic development in SPAs. As the sharing or exchanging of populations and labor within the wider region becomes difficult, SPAs increasingly rely on external and transient sources of human capital. But not only are SPAs more disconnected than other rural geographic units, the Beyond Periphery model suggests that they also become more disconnected over time as a result of improved technology, entrenched external dependencies, and increasingly uneven local development (Taylor et al., 2011; Carson & Carson, 2014), making labor conditions ever more difficult.

2There has been limited empirical examination of the disconnectedness hypothesis for SPAs, mostly through studies looking at migration flows between ‘northern’ cities and their sparsely populated hinterlands in Australia and Sweden (Carson, 2011; Carson et al., 2021). The few available papers suggest that migration and labor systems in northern SPAs differ from those in other rural contexts as they lack stable core-periphery relationships within their regions. In the more densely populated areas in the ‘south’, migration flows to and from rural areas typically follow conventional core-periphery structures and are dominated by the nearest large city (Argent et al., 2008), such that a comparatively high percentage of migrants out of the rural periphery go to the city, and a high percentage of migrants to the periphery come from the city. These percentages are interpreted to be substantially lower in the sparsely populated north for a number of reasons. Regional cities in the north, despite relatively strong economic growth, still have smaller populations than those in the south, and there are fewer intervening city alternatives to attract and retain a large enough labor market able to circulate within the region (Carson et al., 2021; Bjarnason et al., 2021). Potential migrants are also often temporary and have stronger relationships with more distant rather than local origins and destinations, making them less likely to stay and move within the north when their employment status or life stage changes (Taylor et al., 2022; Thurmer et al., 2019).

3None of the papers examining the disconnectedness hypothesis have tested whether local migration intensity, i.e. the percentages of internal migration exchanges within a region, changes over time in either absolute (considering just the northern SPAs) or relative (compared to the south) terms. None of the papers have investigated the impacts of ‘disconnectedness’ or low local migration intensity on labor market development. It remains purely conjecture, therefore, that northern SPAs face labor market challenges (and demographic challenges more generally) precisely because they are disconnected from local labor markets.

4This paper starts by addressing these gaps in validation of the disconnectedness hypothesis by examining the case of northern SPAs in Sweden. Northern Sweden, and particularly the municipalities in the inland which are distant from the regional cities along the coast, contain some of the most sparsely populated areas in Europe (Gløersen, 2012) and have struggled in recent decades to attract and retain key labor, and populations more broadly. The region is characterized by a long history of economic dependence on natural resource industries (forestry, mining, energy, and tourism), persistent outmigration and population decline in the inland, and a simultaneous concentration of people and services in a few urban centres (Hedlund & Lundholm, 2015). Researchers in Sweden have, thus, questioned the extent to which the north is internally disconnected in their economic and demographic development pathways (Eimermann et al., 2022; Carson et al., 2021), and what this could mean for labor and economic development in the inland, particularly as the region appears to become even more polarized in the wake of recent large-scale investments in ‘green transition’ projects (Lundmark et al., 2022).

5This research specifically examines the disconnectedness hypothesis by comparing local migration intensity of working age people in Sweden’s northern SPAs to intensity in other Swedish municipalities over a period of 21 years (2002 – 2023) and by analyzing the relationships between change in intensity, net-migration position and population growth. The expectation is that local migration intensity will be substantially lower in northern SPAs than in other municipalities, that it would have become increasingly lower during the time period, and that low intensity will be at least correlated with low population growth and negative net-migration. The paper proceeds with a short summary of previous research into the disconnectedness hypothesis within the context of the Beyond Periphery model of SPAs. It then describes how ‘disconnectedness’ is operationalized in this research, and how SPAs in Sweden might be identified. The research is then described, and implications for understanding SPAs as a particular geographical unit are discussed.

Background

6The disconnectedness hypothesis is drawn from the Beyond Periphery model of sparsely populated or remote areas, which conceptualises SPAs as sharing eight key characteristics starting with the letter ‘D’ (Carson & Carson, 2014). The model suggests that SPAs are distinct from other rural geographies because they are –

  1. Disconnected from systems which produce the necessary resources for demographic and economic development – people, knowledge, financial capital (Le Tourneau, 2020; Karácsonyi & Taylor, 2023);

  2. Distant from those systems so that efforts to become more connected are costly and likely to have limited impacts (Lundmark et al., 2022);

  3. Dependent, nevertheless, on those distant and disconnected systems to provide the necessary resources which cannot be produced in sufficient quantities internally (Huskey, 2006);

  4. Discontiguous in their development history – meaning that even proximate settlements have different origin stories (as mining towns, agricultural hubs, Indigenous settlements, tourism resorts and so on), limiting the extent to which shared values can be used to underpin collective development (Le Tourneau, 2020);

  5. Diverse in demography, economy and other characteristics such that there is no meaningful ‘average’ experience of development (Taylor et al., 2022);

  6. Detailed in that the actions of even a small number of people (to move in or out, to invest or withdraw investment) can have substantial impacts. Specific events, like the Global Financial Crisis in 2007-08 or the Covid19 pandemic of 2020-22, along with more localized economic boom and bust cycles, can therefore have substantially greater impact on populations in SPAs (Petrov et al., 2021);

  7. Dynamic in that the configurations of people who can create these impacts are regularly changing due to high migration flows of key workers (Vuin, 2019; Eimermann et al., 2022); and

  8. Delicate in their political and social status – on the one hand typically viewed as central to national identity and national ‘frontier’ stories (Klinger et al., 2023; Le Tourneau, 2020), and on the other as sites of social and economic underperformance (Muilu, 2021; Taylor et al., 2011).

7The Beyond Periphery model initially emerged out of dissatisfaction with the concept of the ‘periphery’ (drawn from centre-periphery theory) in explaining the human and economic geographies of SPAs. As critiqued by Pugh and Dubois (2021), the periphery concept has been applied to a multitude of contexts and scales, ranging from global to sub-national regions, from sparsely populated remote territories, to urban fringes and to intra-metropolitan disadvantaged neighborhoods. Not only does this raise problems for distinguishing between different types of peripheries, but more problematically it assumes that a periphery (however defined) is dependent and subordinate to a particular centre on which it relies for access to markets, capital, labor, and political decision-making. The Beyond Periphery model argues that such relationships are much more complicated and unpredictable in SPAs, as distant markets (e.g. for natural resources or tourism) and sources of capital and labor are volatile and do not necessarily align with designated political centres. In this sense, they are not just spatially located beyond what is usually considered the rural or functional hinterland of an urban centre, but they are disconnected from a relational perspective as they lack a stable ‘go to’ centre providing resources for development.

8In short, the Beyond Periphery model, along with recent work by Le Tourneau (2020) and Karácsonyi & Taylor (2023), suggests that SPAs cannot simply be reduced to extreme cases of rural peripheries. There is a need for greater attention to the specific spatial, demographic, economic, social and relational characteristics defining SPAs, which have been variously referred to as fringes, edges, margins, frontiers, along with other country-specific terminologies (Le Tourneau, 2023). While it is outside the scope of this paper to discuss the different nuances and origins of these terms, they have in common that they consider SPAs as being located ‘somewhere beyond’ areas of continuous settlement and exist as transition zones between permanently inhabited and uninhabited spaces, although such zones rarely reflect a clear and sharp outer limit as some of the above terminology implies. Instead, they comprise large and internally heterogenous spaces that may simultaneously contain larger urban centres, dispersed economic growth hotspots (eg. administrative centres, mining towns or tourism resorts), declining settlements that have lost their economic base, pockets of Indigenous and other ethnic or social minorities, and large stretches of uninhabited or only temporarily occupied land. How well such diverse regional conglomerates or mosaics are connected internally as part of an overall sparsely populated region, for example through migration, labor and other mobility exchanges, market linkages or socio-cultural relations, may be an important indicator in understanding and defining what is and what isn’t part of SPAs.

9Ultimately, SPAs are not just expected to be ‘more’ disconnected, dynamic and so on than other rural geographies, but so fundamentally different that, in the terms of this special issue, they are ‘worlds apart’ and require radically new approaches to both address development challenges and exploit innovation opportunities (Pugh & Dubois, 2021). Pugh and Dubois (2021) emphasize that the ‘Ds’ in the Beyond Periphery model reflect the difference between SPAs and other geographical units rather than implying that SPAs are inherently disadvantaged. Such differences could be used to their advantage, for example in developing niche innovations and localized economies through digitalization, smart specialization or ‘attractive mobilities’ such as tourism (Dubois et al., 2017; Lundmark et al., 2022). Typically, however, the disconnectedness of SPAs has been seen as a barrier to development, particularly when it comes to labor development (Strasser, 2016). If SPAs do not have stable core-periphery relationships with their nearest urban centre to access labor, they need to source labor from multiple external and volatile sources instead, which is not only costly but comes with continuous uncertainty about where the next cohort of labor might come from and how it might be prepared for the SPA context. This theme of disconnectedness as a barrier to labor development in SPAs has been examined in the context of SPAs’ access to university qualified labor, and specifically labor required in sectors such as health, tourism and the recent green industry transition (Carson et al., 2021; Lundmark, 2006; Lundmark et al., 2022), primarily in northern Australia and northern Sweden.

10Possible explanations for disconnectedness may come from migration theories, which have been shown to apply in particular ways to SPAs (Thurmer et al., 2019). Migration patterns involving SPAs as origins or destinations are claimed to be inherently unpredictable and unstable. While the fundamentals of out-migration (‘stepping up’ to larger urban centres, pursuing academic and career development) and in-migration (‘stepping down’ for lifestyle or amenity reasons, ‘escalating’ careers, returning to childhood environments) remain (Rees & Sander, 2019), migrants from SPAs are less constrained by internal or ‘local’ core-periphery migration systems (Carson et al., 2021). The universities and jobs they seek are likely to be extra-local (Corcoran et al., 2010), nearby urban centres are likely to be comparatively small with limited pulling power (Bjarnason et al., 2021), necessitating longer journeys to ‘step up’, which consequently opens up a larger pool of potential destinations. Should migrants seek to return to their SPA origins, they return from a wider range of places than might be expected in a more standard core-periphery migration model (Huskey et al., 2004). In-migrants stepping down in search of lifestyle or career changes, once they have decided that they are not constrained by proximity to their previous residence, have multiple intervening destination options that appear similarly attractive (Axelsson & Hedberg, 2025; Eimermann et al., 2022).

11Parallels in experiences of SPA migration, while not specifically mapped to the Beyond Periphery model or the disconnectedness hypothesis, can be seen in research emanating from Canada, Alaska, and the European Arctic (Howe et al., 2014, Hillier et al., 2020, Heleniak, 2021, Bjarnason et al., 2021). Many of these suggest that in-migration is likely to be more ‘messy’ than out-migration as pathways out of SPAs are relatively well established (Eimermann et al., 2022). Within working age populations, there are likely to be differences in disconnectedness among age groups (and gender and other factors not considered in this study). Youth and younger adults are likely to be highly mobile and open to a wider range of destinations (and hence origins when migrating to SPAs), but also more influenced by local migration pathways which are familiar and accessible (Rönnlund & Tollefsen, 2024). Those in mid-career and family formation stages tend to have narrower migration paths as they balance work and family responsibilities (Sandow & Lundholm, 2023), while there is some evidence that migration among older age workers may be driven in part by escalator motives (related to better pay and faster career progression) similar to those of younger adults (Dyrting et al., 2020).

12The Beyond Periphery model also asserted that, not only are SPAs disconnected from human and capital resource flows, but they tend to become increasingly disconnected over time. Examples from Outback Australia specifically referred to transport and communications initiatives which caused increasing disconnection not in spite of efforts to address the challenges of remoteness, but because of them (Cleary, 2012). Initiatives to improve the quality of flying doctor services, for example, have resulted in reducing the range of places where increasingly sophisticated aircraft can land. Similar claims have been made in relation to distance-bridging technologies in remote service delivery (e.g. remote health consultations, distance education, online retail or banking), which essentially normalizes the ongoing withdrawal of physical services and staff from SPAs into larger urban centres (Jonsson et al., 2021) and reinforces social isolation (Le Tourneau, 2023). Coupled with increasing concentration of resources even within SPAs, this leads to inequality and barriers to participating in ‘local’ resource flows (Ryser et al., 2019; Taylor et al., 2011).

13Despite increasing conceptual work on disconnectedness, there is still very little empirical evidence for the hypothesis that SPAs are disconnected from labor markets and other resource flows, and there has been no empirical investigation of the hypothesis that SPAs become more disconnected over time. Carson’s (2011) paper showed that migration flows from the capital city to the remainder of Australia’s Northern Territory (its most sparsely populated jurisdiction) were less than two thirds the expected volume when compared to flows in other jurisdictions. However, there was no attempt to examine how these flows might have changed over time, or to assess how the statistically significant difference in flows for the Northern Territory might be functionally important in terms of the cost of labor development. Carson et al. (2021) looked at the migration of university qualified labor within seven city-hinterland regions in northern Australia and northern Sweden. Their main conclusion was that there was great diversity in regional migration patterns between the seven regions, with no universal experience of disconnectedness, but a general tendency towards higher levels of disconnectedness in the Australian cases. Again, there was no attempt to compare SPA experiences with those of other rural geographies or changes over time. Drawing on population and migration data from Sweden, this research extends previous work on disconnectedness in SPAs by adding a more longitudinal perspective. Specifically, we compare local migration intensities for northern SPAs and other Swedish municipalities, considering differences across age groups, and we analyze the relationships between change in migration intensity, net-migration position, and population growth over time.

Methods

Defining the study area

14One of the challenges in examining the disconnectedness hypothesis for SPAs is the multitude of scales at which SPAs are defined and investigated. Scales might refer to supra-jurisdictional as single entities, for example when considering all of ‘desert’ Australia or Europe’s declared northern sparsely populated areas, including all of northern Sweden (Gløersen, 2012). They may also refer to single jurisdictions like Australia’s Northern Territory or Sweden’s Norrbotten county – including their densely populated cities (Pugh & Dubois, 2021), or to smaller and more localized scales, such as municipalities and even sub-municipal areas which are distinguished from municipal administrative centres (Eimermann et al., 2022). Attempting better definitional clarity is beyond the scope of this paper except to emphasize that how and why SPAs have been defined in a particular way needs to be explicit in any empirical study.

15While Sweden has a formal geographical classification of municipalities, it does not identify SPAs explicitly (Sveriges Kommuner och Regioner, 2023). Rather, it has four broad categories of municipalities based on size of the largest internal urban centre and distance to the nearest urban centre of more than 15 000 residents. The first category includes Sweden’s largest three cities – Stockholm, Göteborg and Malmö – and the municipalities from which workers commute daily to those cities. In the second level, there are 23 cities with more than 50 000 residents, and nearly 100 municipalities in daily commuting distance. The third level breaks down into 27 small cities of between 15 000 and 50 000 residents and 51 municipalities commuting distance from those, and, finally within the fourth level, 56 municipalities which neither have even a ‘small city’ nor are within commuting distance of such a city (‘landsbygdskommuner’). Eighteen of these are within three of Sweden’s four northern counties commonly referred to as Norrland, and all of these are in the western inland parts of those counties. They all had population densities of less than two persons per square kilometre in 2023 (thus meeting one threshold proposed by Le Tourneau and Vaguet (2024) in the call for this issue), and only one other municipality in Sweden (the mountain municipality Älvdalen in Dalarna County, immediately south of Norrland) had a population density below two. For this study the eighteen northern, inland and sparsely populated municipalities are considered to be Sweden’s north SPA (Figure 1). Consequently, SPAs here are defined by population density (less than two persons per square kilometre), population size (no more than 15 000 residents in any urban area), and distance (outside of daily commuting distance of any larger urban centre). The Swedish classification also presents a logical comparison area comprising the 31 landsbygdskommuner in southern Sweden (the rural south) which meet population size and distance criteria, but have higher population densities than north SPAs.

Figure 1: Sweden’s north SPAs

Figure 1: Sweden’s north SPAs

Defining ‘disconnectedness’ and the time series

16Previous research has measured (dis)connectedness as a function of migration exchange, typically between the largest city in a sparsely populated region and other geographic units (municipalities, sub-regions) of that region (Carson et al., 2021; Bjarnason et al., 2021; Argent et al., 2008). In this research, connectedness is conceptualized as not just the strength of relationship between geographic units and one central place, but among all geographic units within a jurisdiction. The measures of (dis)connectedness are the proportion of all in-migration to a municipality that comes from other municipalities within the same county (local in-migration intensity) and the proportion of all out-migrations from a municipality that go to another municipality within the same county (local out-migration intensity).

17The time series used in the research (2002-23) starts with the first-year data on local migration intensity was published by Statistics Sweden and ends with the most recently available data at the time the research was conducted. Tests of statistical significance (specifically t-tests) were conducted, but there is the persistent challenge of deciding whether findings which are statistically significant are also functionally important when trying to define SPAs as ‘worlds apart’ from other geographies. Likewise, where definitive statistical measures of significance are unavailable (in comparing magnitudes or regression coefficients, for example), one needs to decide what differences between SPAs and other geographies or what differences between time periods are likely to be important.

18Overall, for a number of reasons the conclusions of this research regarding the applicability of the disconnectedness hypothesis to SPAs in Sweden may be questioned:

  1. This research considers disconnectedness in isolation from other defining dimensions of SPAs, while there is the possibility that SPAs are defined by a combination of multiple (known and unknown) dimensions.

  2. This particular operationalization of SPAs (municipal level, population density, size and distance qualifiers) may be inappropriate.

  3. This particular operationalization of the disconnectedness concept may not adequately represent that concept.

  4. The comparison units (rest of Sweden and non-sparsely populated rural municipalities or ‘landsbygdskommuner’) may be inappropriate for such study.

  5. The time series may be insufficiently long or short, or commences or ends at inappropriate times to make both point in time and time series analysis valid.

19These challenges are recognized and commented on in the Discussion, with the operationalization reflecting both theoretical considerations (that ‘connectedness’ is to more than a single urban unit) and practical constraints (availability of data).

Data extraction and analysis

20Data were drawn from Statistics Sweden Statistical Database (https://www.statistikdatabasen.scb.se/​pxweb/​en/​ssd/​) on 29th October 2024. Extracted tables were –

  • Population by region, marital status, age and sex. Year 1968-2023

    • For each municipality in Sweden

    • No specification of marital status

    • Separately for ages 15-24 (youth), 25-34 (early career), 35-54 (mid career), 55-64 (late career) and for the total population aged 15-64.

    • No specification for sex

    • Years 2002-2023

  • Population density per sq. km, population and land area by region and sex. Year 1991-2023

    • Population density per sq. km

    • For each municipality in Sweden

    • No specification for sex

    • Year 2023

  • Migration by region, age and sex. Year 1997-2023

    • In-migrations, out-migrations, immigrations and emigrations

    • For each municipality in Sweden

    • Separately for ages 15-24, 25-34, 35-54, 55-64 and for the total population

    • No specification for sex

    • Years 2002-2023

  • Migration by region of birth, region, age and sex. Year 2002-2023

    • In-migrations from other municipalities within the county, Out-migrations to other municipalities within the county

    • For each municipality in Sweden

    • Separately for ages 15-24, 25-34, 35-54, 55-64 and for the total population

    • No specification for sex

    • No specification for region of birth (born in Sweden or foreign-born)

    • Years 2002-2023

21Separate spreadsheets were created in Microsoft Excel for each age group and each year, containing –

  • Municipality code and municipality name

  • Population

  • Population density per sq. km

  • Total in-migrations (inmigrations + immigrations)

  • Total out-migrations (outmigrations + emigrations)

  • In-migrations from other municipalities within the county

  • Out-migrations to other municipalities within the county

22Derived variables representing local (within county) in-migration intensity and local out-migration intensity were added to these spreadsheets –

  1. Proportion of in-migrations which were from municipalities within the county (total in-migrations / in-migrations from other municipalities within the county (Pin))

  2. Proportion of out-migrations which were to municipalities within the county (total out-migrations / out-migrations from other municipalities within the county (Pout))

23Municipalities were then classified geographically using the Classification of Swedish Municipalities (Sveriges Kommuner och Regioner, 2023) –

  1. Population density less than 2.0 per square kilometre and classified as rural municipalities (‘landsbygdskommun’) or rural municipalities with tourism industry (‘landsbygdskommun med besöksnäring’): north SPA – 18 municipalities

  2. Population density greater than 2.0 per square kilometre and classified as rural municipalities or rural municipalities with tourism industry: rural south – 31 municipalities with consistent data across the time series

  3. Rest of Sweden – the remaining 240 municipalities

24Statistical analysis was performed in the BlueSky Statistics open-source software (https://www.blueskystatistics.com/​). T-tests (for independent samples) were conducted comparing mean Pin and Pout for north SPA and the rest of Sweden and north SPA and rural south for each age group and total population for the years 2002 and 2023. T-statistic values were considered significant if the probability (two tailed, unequal variances) of obtaining that value (p-value) was less than 0.05 (*), less than 0.01 (**) or less than 0.001 (***).

25Time series analysis was conducted in Microsoft Excel. The raw differences in local in-migration intensity between north SPA and the rest of Sweden and between north SPA and rural south were calculated for 2023 and 2002. The change in those differences (difference 2023 – difference 2002) was calculated. Changes in difference of more than 7% were considered potentially important because that magnitude of change was typically assessed as statistically significant (**) in the t-tests.

26Each series for north SPA (Pin x age and Pout x age) was assessed for –

  1. Change in intensity 2023-2002,

  2. Range of intensity (highest value – lowest value) 2002-23,

  3. Variance of series (measured by the coefficient of variation – standard deviation of the range / mean of the range). A coefficient of variation greater than 0.50 was considered substantial,

  4. Linearity of the series (measured by the regression coefficient (r2) plotting the series against the year range). A series was considered to have a linear trend if r2 was greater than 0.40 (moderate), greater than 0.60 (strong) or greater than 0.80 (very strong)

27Each series was then assessed to identify outliers using an interquartile range (IQR) technique (the range between the highest score in the third quartile and the highest score in the first quartile). If the value for any year exceeded the IQR * 1.5, that year was considered an outlier.

28Finally, each series was plotted on a line graph and visually assessed for apparent clusters of high and low values (even if not exceeding IQR*1.5) and sub-trends (sections of the graph including more than 9 years where there was a linear trend).

Results

In- and out-migration intensity, 2023

29Local migration was important, with 30% of migrants into northern sparsely populated municipalities (north SPA) arriving from the same county in 2023 (Table 1). While this was a significantly lower proportion than for the rest of Sweden (45%), it was similar to that for the southern landsbygd municipalities (the rural south). The difference in local in-migration intensity between north SPA and the rest of Sweden was smaller for youth (7% compared with 15% for the total population) but still statistically significant. Youth in-migration intensity was similar in north SPA and the rural south. In contrast, local in-migration intensity among early and mid-career populations was significantly lower in north SPA than in the rural south, although actual differences (9% and 6%) were not as substantial as those between north SPA and the rest of Sweden (18% and 19%). In-migration intensity for late career populations was significantly lower in north SPA than the rest of Sweden, but once again similar in north SPA as in the rural south.

Table 1: Proportion of in-migrants to the same county (local in-migration intensity) comparing north SPA, the rest of Sweden and rural south, 2023

 

In-migration intensity

Difference to north SPA

T

P

Total population

North SPA

30%

 

 

 

Rest of Sweden

45%***

15%

5.840

0.000

Rural south

34%

4%

1.508

0.070

Youth

North SPA

38%

 

 

 

Rest of Sweden

45%*

7%

1.914

0.035

Rural south

38%

0%

0.066

0.474

Early career age

North SPA

28%

 

 

 

Rest of Sweden

46%***

18%

6.701

0.000

Rural south

37%*

9%

2.885

0.003

Mid career age

North SPA

25%

 

 

 

Rest of Sweden

44%***

19%

7.526

0.000

Rural south

31%*

6%

1.972

0.028

Late career age

North SPA

32%

 

 

 

Rest of Sweden

49%***

17%

4.473

0.000

Rural south

34%

2%

0.521

0.303

30Local out-migration intensity for the total population, and early, mid and late career populations followed a similar pattern, being significantly lower in north SPA compared to the rest of Sweden, but similar in north SPA as in the rural south (Table 2). Differences between north SPA and the rest of Sweden were small (8-12 %) when compared with differences in in-migration intensity (15-20 %). There were no differences in out-migration intensity among the three geographic units for the youth population.

Table 2: Proportion of out-migrants to the same county (local out-migration intensity) comparing north SPA, the rest of Sweden and rural south, 2023

 

Out-migration intensity

Difference to north SPA

T

P

Total population

North SPA

40%

 

 

 

Rest of Sweden

48%**

8%

2.591

0.008

Rural south

39%

1%

0.327

0.373

Youth

North SPA

44%

 

 

 

Rest of Sweden

49%

5%

1.407

0.087

Rural south

43%

1%

0.274

0.393

Early career age

North SPA

41%

 

 

 

Rest of Sweden

49%**

8%

2.558

0.009

Rural south

39%

2%

0.615

0.272

Mid career age

North SPA

36%

 

 

 

Rest of Sweden

47**

11%

3.459

0.001

Rural south

37%

1%

0.450

0.328

Late career age

North SPA

40%

 

 

 

Rest of Sweden

52%*

12%

2.529

0.010

Rural south

43%

3%

0.495

0.312

31In-migration intensity in north SPA was lower for mid-career populations (25%) than for youth (38%) (t=3.184, p=0.002). In and out-migration intensity in north SPA were otherwise similar for all populations. Out-migration intensity was higher than in-migration intensity for the total population (40% compared with 30%, t=2.680, p=0.006). Out-migration intensity was higher than in-migration intensity for mid-career populations (36% compared with 25%, t=2.944, p=0.003). In- and out-migration intensities for other populations were similar to one another.

Changes in local migration intensity 2002-2023

32There were no significant differences in in-migration or out-migration intensity for north SPA when comparing 2002 and 2023. In relative terms however (comparing north SPA with the rest of Sweden and with the rural south), north SPA had lower in-migration intensity in 2023 compared with 2002. Table 3 shows the in-migration intensity for 2023 and 2002, the differences between north SPA and other geographies in each of those years and the change in magnitude of difference between those two years. For example, the rural south had an in-migration intensity rate 3% lower than that of north SPA in 2002, but 4% higher in 2023, meaning a change in the difference between the two populations of 7% over that time. While the statistical significance of this increase cannot be assessed, Tables 3 and 4 suggest that differences of 7-10% might be substantial. On that basis, changes in differences between rates are likely to be important for all populations except mid-career age and for the rest of Sweden and the rural south. The largest change in that matter was between north SPA and the rural south for the early career age population (an increase of 13%). All changes suggest that north SPA had comparatively lower local in-migration intensity at the end of the period than at the start.

Table 3: Comparison of differences in local in-migration intensity between north SPA, rest of Sweden and rural south 2002-2023

 

In-migration intensity

In-migration intensity

Difference to north SPA

Change in difference

 

2002

2023

2002

2023

2002-2023

Total Population

North SPA

34%

30%

 

 

 

Rest of Sweden

43%

45%

9%

15%

6%

Rural south

31%

34%

-3%

4%

7%

Youth

North SPA

36%

38%

 

 

 

Rest of Sweden

43%

45%

9%

15%

6%

Rural south

35%

38%

1%

8%

7%

Early career age

North SPA

33%

28%

 

 

 

Rest of Sweden

42%

46%

8%

16%

8%

Rural south

28%

37%

-6%

7%

13%

Mid career age

North SPA

33%

25%

 

 

 

Rest of Sweden

44%

44%

10%

14%

4%

Rural south

3%

31%

-4%

1%

5%

Late career age

North SPA

36%

32%

 

 

 

Rest of Sweden

45%

49%

11%

19%

8%

Rural south

29%

34%

-5%

4%

9%

33When it comes to local out-migration intensity, differences in rates between north SPA and the other two geographic units remained similar (Table 4), with the exception of out-migration intensity for north SPA youth that was relatively lower than for youth in the rural south at the end of the period.

Table 4: Comparison of differences in local out-migration intensity between north SPA, rest of Sweden, and rural south 2002-2023

 

Out-migration intensity

Out-migration intensity

Difference to north SPA

Change in difference

 

2002

2023

2002

2023

2002-2023

Total population

North SPA

42%

40%

 

 

 

Rest of Sweden

48%

48%

6%

8%

2%

Rural south

39%

39%

-3%

-1%

2%

Youth

North SPA

42%

44%

 

 

 

Rest of Sweden

47%

49%

5%

9%

4%

Rural south

38%

43%

-4%

3%

7%

Early career age

North SPA

40%

41%

 

 

 

Rest of Sweden

47%

49%

5%

9%

4%

Rural south

36%

39%

-6%

-1%

5%

Mid career age

North SPA

41%

36%

 

 

 

Rest of Sweden

50%

47%

8%

7%

-1%

Rural south

40%

37%

-2%

-3%

-1%

Late career age

North SPA

41%

40%

 

 

 

Rest of Sweden

50%

52%

8%

12%

4%

Rural south

41%

43%

-1%

3%

4%

34Figure 2 plots changes in in-migration intensity for north SPA across the time period, with the 2002 rate as the baseline. While interquartile ratio analysis revealed no outlier years there were years of low intensity (except for late career age) in 2016 and 2017, and relatively high intensity in 2019 and 2020 (including for late career age). While over the entire time period there was no linear trend of changes in in-migration intensity, there was a strong linear trend of decline in intensity between 2002 and 2018 (RSQ=0.78 for total population). The trend also existed for youth (RSQ=0.76) and mid-career age (0.65), but not for early career age (0.37) or late career age (0.01).

Figure 2: Changes in in-migration intensity, 2002-2023 (baseline 2002)

Figure 2: Changes in in-migration intensity, 2002-2023 (baseline 2002)

35Analysis of changes in out-migration intensity did not reveal any outlier years, clusters or sub-period linear trends (Figure 3).

Figure 3: Changes in out-migration intensity, 2002-2023 (baseline 2002)

Figure 3: Changes in out-migration intensity, 2002-2023 (baseline 2002)

Local migration intensity and population change

36Increases in in-migration intensity for north SPA were strongly negatively associated with decreases in the number of in-migrants, the number of out-migrants and net migration for the total population and for mid-career age population (Table 5). Increases in in-migration intensity were also associated with decreases in net migration for youth and early career age, and increases in the youth population, despite being associated with decreases in in-migrant numbers and net migration, reflecting the ‘topping up’ of the youth population each year through natural increase as the comparatively large 14-year-old population turns 15 years old and the smaller 24-year-old cohort ages out. This effect was particularly notable in the early part of the period (to 2007-8). Following that, there was no association between youth in-migration intensity and youth population growth (r2 = 0.29).

Table 5: Relationships (r2) between in-migration intensity and population change, north SPA 2002-23

 

Population growth

Nr. of in-migrants

Nr. of out-migrants

Net migration

Total Population

0.35

0.88↓

0.74↓

0.81↓

Youth

0.47↑

0.76↓

0.00

0.79↓

Early career age

0.32

0.26

0.02

0.46↓

Mid career age

0.11

0.83↓

0.51↓

0.69↓

Late career age

0.06

0.02

0.05

0.02

37Table 6 shows fewer relationships between local out-migration intensity and population development. Increase in out-migration intensity was associated with decrease in number of in-migrants and number of out-migrants for the total population and for mid-career age, but the impact on both these measures was insufficient to have an effect on net migration.

Table 6: Relationships (r2) between out-migration intensity and population change, north SPA 2002-23

 

Population growth

Nr. of in-migrants

Nr. of out-migrants

Net migration

Total Population

0.11

0.42↓

0.41↓

0.35

Youth

0.03

0.16

0.07

0.06

Early career age

0.00

0.07

0.30

0.00

Mid career age

0.06

0.47↓

0.45↓

0.28

Late career age

0.03

0.07

0.04

0.00

Discussion

38Sweden’s northern sparsely populated municipalities (north SPA) were clearly more disconnected than the rest of Sweden based on the measure of local (within-county) migration intensity. The difference in connectedness was starker when examining in-migration intensity than out-migration intensity. However, north SPA were not consistently more disconnected than the rural south (landsbygd municipalities). The differences between in-migration and out-migration connectedness might arise from different migration processes underpinning them. Out-migration might more consistently involve step migration with migrants first moving to relatively proximate urban centres and then later onwards to more distant places (Howe et al., 2014). In-migration, on the other hand, might be more influenced by escalator migration (a less predictable pattern) than stepping down the urban hierarchy. This might particularly apply to youth, who step through education and early career opportunities (Dyrting et al., 2020). This is reflected in the relatively small differences in local out-migration intensity for youth among the three geographic units included in this study. In-migration in older age groups might also include an element of escalator migration along with amenity migration, often with diverse origins (Huskey et al., 2004; Eimermann et al., 2022), and return migration where the ‘return’ might be from places several steps away from the original out-migration destination (Sandow & Lundholm, 2023). Whatever the validity of these conjectures, there is evidence that local out-migration intensity is more stable, and potentially more predictable in terms of destinations, than in-migration.

39Table 7 summarises some of the key findings presented in this Discussion. Statistically significant differences in migration intensity are identified with upwards arrows (indicating that intensity in rest of Sweden or rural south was higher than in north SPA).

Table 7: Summary of key findings

Variable

Population

North SPA

Rest of Sweden

Rural South

In-migration intensity

Total

30%

45% (↑)

34%

Out-migration intensity

Total

40%

48% (↑)

39%

In-migration intensity

Youth

38%

45% (↑)

38%

Out-migration intensity

Youth

44%

49%

43%

In-migration intensity

Early career

28%

46% (↑)

37% (↑)

In-migration intensity

Mid-career

25%

44% (↑)

37% (↑)

% increase in in-migration intensity 2002-23

Total

-12%

5%

10%

% increase in out-migration intensity 2002-23

Total

-5%

0%

0%

% increase in in-migration intensity 2002-23

Early career

-15%

10%

32%

Absolute increase in in-migration intensity 2002-2023

Total

-4%

2%

3%

Absolute increase in in-migration intensity 2002-2018

Total

-8%

-3%

-2%

40That disconnectedness is often similar for the rural south as for north SPA raises interesting questions about the definition of sparsely populated areas, at least in a Swedish context. The rural south is distinguished from the north SPA by its population density (no municipalities have fewer than 2 persons per square kilometre), population size (municipalities are typically double the size of north SPA municipalities), and the size of close larger urban centres which are nevertheless similarly distant as northern cities from north SPA. In addition, municipalities in the rural south have multiple large urban centres similarly distant to one another. These differences might slightly matter when it comes to disconnectedness (in-migration for early and mid-career age populations), but otherwise it is difficult to distinguish the rural south and north SPA on the measures used here.

41The relatively short timeframe for this data series (roughly two decades) makes it difficult to be definitive about whether disconnectedness of the north SPA has increased over time. Again, the evidence that this has occurred is stronger for in-migration than out-migration. There may have been higher increases in relative in-migration intensity between north SPA and the rural south than between north SPA and the rest of Sweden – most prominent for the early career age population.

42While the north SPA may have become relatively more disconnected over time, there is no evidence that it became more disconnected in absolute terms. However, and again for in-migration rather than out-migration, had the data set for this study ended in 2018 or even 2019, the conclusion would have been that the north SPA had become substantially more disconnected over time. Something happened around this time to (at least temporarily) reverse this trend. Perhaps this is related to the wave of refugee in-migrants arriving in 2014-16 and decreasing the proportion of within-county in-migrants for those years, followed by the Covid crisis reducing the propensity for people to migrate long distances (and particularly across international borders). Whatever the case, the research has at least revealed the potential for major events to disrupt relationships between SPAs and their migration partners (Petrov et al., 2021). However, other notable events on both global (e.g. the Global Financial Crisis in 2007-08) and regional or even local scales (e.g. the urban expansion surrounding the 2014 European Capital of Culture in Umeå or the recent large-scale investments connected to green transition projects starting in 2020) do not appear to be reflected in the current data. Nevertheless, there are hints that something may have impacted youth in-migration patterns in 2006-07 and late career age patterns around 2011-12.

43Previous research has postulated that disconnectedness from a migration perspective matters because it makes it difficult to identify potential sources of migrants and therefore increases the costs of recruiting labor (Carson, 2011). This may be true, but this research suggests that increased connectedness (higher local migration intensity) leads to lower in-migration rates for most age groups and poorer net-migration position for all except the late career age population. Local labor markets are so considerably smaller than distant ones that, even though it may be less resource intensive to tap them, for example through participation in higher education and employment pathways (Rönnlund & Tollefsen, 2024), their potential to service labor needs is low.

Conclusions

44In summary, the research provides some support for the suggestion that SPAs are more disconnected than other geographic units, and that they become more disconnected over time. However, this case study is by no means as conclusive as previous works by Carson and colleagues suggest (Carson, 2011; Carson et al., 2021), largely inspired by their northern Australian observations. Compared to other rural areas in Sweden, the differences are often negligible. It would be difficult to argue that the north SPA is ‘worlds apart’ from the rural south, although it probably is ‘worlds apart’ from urban settings. This raises specific and general questions for research into SPAs. In specific terms, one question is whether the inland north of Sweden should be considered a sparsely populated area at all. It meets the (perhaps arbitrary) requirement of having fewer than two inhabitants per square kilometer, but other parameters around population size and distance from larger centres may also be required to identify ‘true’ SPAs. An alternative angle for this discussion is whether the rural south actually qualifies as a sparsely populated area, at least when it comes to challenges in accessing and mobilizing local labor, despite having a relatively high population density and relatively large municipal populations. The comparable distances from larger urban centres may be a definitive factor. Whatever the case, a clear distinction between SPAs and other rural geographies seems less straightforward in Sweden (and much of northern Europe) than in other countries featuring in the growing SPA literature (e.g. Australia, Canada, the US, Brazil etc). This may be due to a combination of factors, including different colonial histories and settlement patterns, different spatial planning traditions, the continuing legacies of the Nordic welfare state model, and a stronger sense of shared regional identities and mobility practices, making northern SPAs in Sweden less disconnected from a migration perspective than SPAs in more remote and isolated ‘edges’ facing more polarized socio-economic development (Taylor et al., 2011; Karácsonyi & Taylor, 2023).

45More generally, if the northern inland of Sweden should continue to be identified as a sparsely populated area, and as such ‘worlds apart’ from other geographic units, then the validity of the disconnectedness concept as an independently defining feature of SPAs needs to be questioned. Rather, disconnectedness may be one of a set of features, such as in the initial Beyond Periphery model comprising eight ‘D’ characteristics (Carson & Carson, 2014), which act together to define SPAs. Apparently small differences in (dis)connectedness may become critical when coupled with other factors like distance from markets, diversity of economic, social and demographic characteristics, and so on. Empirical research into SPAs needs to account for the combination of these rather than consider each in isolation.

46While limited by being a single factor study, this research nonetheless contributes to the broader discussion about SPAs as a particular geography (Le Tourneau, 2020; Karácsonyi & Taylor, 2023). It has explored how SPAs compare with other types of ‘rural’. It has demonstrated how SPAs are likely to ‘behave’ differently for different populations (here based on age). It has suggested the potential for global scale events like the 2014-16 refugee crisis and the Covid19 pandemic to impact demographic trends in SPAs, raising the question of whether more local events can also be impactful. More importantly, it has broadened the conversation about what being a sparsely populated area means when pursuing development goals, in this case labor development. It is often tempting to portray SPAs as disadvantaged. Rather, they should be considered as different (even if not ‘worlds apart’) (Pugh & Dubois, 2021). Managing difference - in this case by pursuing more resource intensive externally-oriented labor recruitment strategies - may be difficult, but may also have more desirable outcomes than imposing assumptions associated with traditional core-periphery models where labor is more easily and sustainably recruited from close-by urban centres (Strasser, 2016).

47Overall, we want to redirect attention to the social relationships and human capital exchanges within SPAs as a key feature underpinning their fragile human and economic geographies. Rather than trying to define and identify SPAs as absolute spaces based on universal spatial, demographic and socio-economic indicators, there is a need to consider how SPAs evolve as relational spaces over time, and this includes consideration of the changing relationships with both external and internal sources of human capital. While this research has been limited to migration as one particular form of internal human capital exchange, more empirical research is needed to better understand the nature and strengths of different social, economic and cultural relations that may contribute to (dis)connectedness in SPAs, how such relations change over time, and what the implications are for local development. It is still unclear whether striving for increased connectedness (for example, through increased transport and mobility connections, stronger local market linkages, the transfer of locally-produced skills, and region-building activities aimed at generating a stronger sense of cohesion) would help address common development challenges in SPAs. Stronger connections within the region, for example through internal core-periphery relationships, may ultimately benefit the few larger urban centres at the expense of the hinterland (Argent et al., 2008; Lundmark et al., 2022), while leaving smaller settlements without alternative pipelines of human capital.

Haut de page

Bibliographie

Argent Neil, Rolley Fran, Walmsley Jim, 2008, “The Sponge City Hypothesis: does it hold water?,” Australian Geographer, 39(2), pp. 109–130.

Axelsson Linn, Hedberg Charlotta, 2025, “Placing Migration: Towards More‐Than‐Relational Geographical Migration Studies,” Population, Space and Place, 31(2) : e2872.

Bjarnason Thoroddur, Stockdale Aileen, Shuttleworth Ian, Eimermann Marco, Shucksmith Mark, 2021, “At the intersection of urbanisation and counterurbanisation in rural space: Microurbanisation in Northern Iceland,” Journal of Rural Studies, 87, pp. 404–414.

Carson Doris A., Carson Dean B., Lundström Linus, 2021, “Northern cities and urban–rural migration of university‐qualified labour in Australia and Sweden: Spillovers, sponges, or disconnected city–hinterland geographies?,” Geographical Research, 59(3), pp. 424–438.

Carson Dean B., 2011, “Skilled Labour Migration Flows to Australia’s Northern Territory 2001-2006: Beyond Periphery?,” The Australian Journal of Labour Economics, 14(1), pp. 15–33.

Carson Dean B., Carson Doris A., 2014, “Local economies of mobility in sparsely populated areas: Cases from Australia’s spine,” Journal of Rural Studies, 36, pp. 340–349.

Cleary Jennifer, 2012, “Business Exchanges in the Australian Desert: It’s About More Than the Money,” Journal of Rural and Community Development, 7(1), pp. 1–15.

Corcoran Jonathan, Faggian Alessandra, Mccann Philip, 2010, “Human Capital in Remote and Rural Australia: The Role of Graduate Migration,” Growth and Change, 41(2), pp. 192–220.

Dubois Alexandre, Kristensen Iryna, Teräs Jukka, 2017, “Outsmarting geography: implementing territorial innovation strategies in sparsely populated regions,” European Planning Studies, 25(8), pp. 1316–1333.

Dyrting Sigurd, Taylor Andrew, Shalley Fiona, 2020, “A life-stage approach for understanding population retention in sparsely populated areas,” Journal of Rural Studies, 80, pp. 439–451.

Eimermann Marco, Adjei Evans Korang, Bjarnason Thoroddur, Lundmark Linda, 2022, “Exploring population redistribution at sub-municipal levels – Microurbanisation and messy migration in Sweden’s high North,” Journal of Rural Studies, 90, pp. 93–103.

Gløersen Erik, 2012, “Renewing the Theory and Practice of European Applied Territorial Research on Mountains, Islands and Sparsely Populated Areas,” Regional Studies, 46(4), pp. 443–457.

Hedlund Martin, Lundholm Emma, 2015, “Restructuring of rural Sweden – Employment transition and out-migration of three cohorts born 1945–1980,” Journal of Rural Studies, 42, pp. 123–132.

Heleniak Timothy, 2021, “The future of the Arctic populations,” Polar Geography, 44(2), pp. 136–152.

Hillier Cathlene, Sano Yujiro, Zarifa David, Haan Michael, 2020, “Will They Stay or Will They Go? Examining the Brain Drain in Canada’s Provincial North,” Canadian Review of Sociology/Revue canadienne de sociologie, 57(2), pp. 174–196.

Howe E. Lance, Huskey Lee, Berman Matthew D., 2014, “Migration in Arctic Alaska: Empirical evidence of the stepping stones hypothesis,” Migration Studies, 2(1), pp. 97–123.

Huskey Lee, 2006, “Limits to growth: remote regions, remote institutions,” The Annals of Regional Science, 40(1), pp. 147–155.

Huskey Lee, Berman Matthew, Hill Alexandra, 2004, “Leaving home, returning home: Migration as a labor market choice for Alaska Natives,” The Annals of Regional Science, 38(1), pp. 75–92.

Jonsson Frida, Carson Dean B., Goicolea Isabel, Hurtig Anna-Karin, 2021, “Strengthening Community Health Systems Through Novel eHealth Initiatives? Commencing a Realist Study of the Virtual Health Rooms in Rural Northern Sweden,” International Journal of Health Policy and Management, p. 1.

Karàcsonyi David, Taylor Andrew, 2023, “Understanding demographic and economic patterns in sparsely populated areas – a global typology approach,” Geografiska Annaler: Series B, Human Geography, 105(3), pp. 228–247.

Klinger Julie, Bennett Mia, Adams Ria-Maria, Armstrong Eleanor, 2024, “The crowded and empty Arctic: Examining research practices, spatial imaginaries, and infrastructures in Kiruna, Sweden,” Lychnos, pp. 57-72.

Le Tourneau François-Michel, 2020, “Sparsely populated regions as a specific geographical environment,” Journal of Rural Studies, 75, pp. 70–79.

Le Tourneau François-Michel, 2023, “Périphéries, franges, marges ou confins ? Les « grands espaces » et leurs populations dans le monde contemporain,” Espace populations sociétés, 2023/1.

Le Tourneau François-Michel, Vaguet Yvette, 2024, “Sparsely populated areas: worlds apart?,” Espace populations sociétés. Available at: https://journals.openedition.org/eps/14298 (accessed on 15 April 2025)

Lundmark Linda, 2006, Restructuring and employment change in sparsely populated areas: Examples from northern Sweden and Finland, Doctoral dissertation, Umeå, Umeå University.

Lundmark Linda, Carson Doris A., Eimermann Marco, 2023, “Spillover, sponge or something else? Dismantling expectations for rural development resulting from giga-investments in Northern Sweden,” Fennia - International Journal of Geography, 200(2).

Muilu Toivo, 2021, “Rural Policies for Sparsely Populated Areas in Finland – Old Problems, New Challenges and Future Opportunities,” European Countryside, 13(2), pp. 479–491.

Petrov Andrey N., Welford Mark, Golosov Nikolay, DeGroote John, Devlin Michele, Degai Tatiana, Savelyev Alexander, 2021, “The ‘second wave’ of the COVID-19 pandemic in the Arctic: regional and temporal dynamics,” International Journal of Circumpolar Health, 80(1): 1925446.

Pugh Rhiannon, Dubois Alexandre, 2021, “Peripheries within economic geography: Four ‘problems’ and the road ahead of us,” Journal of Rural Studies, 87, pp. 267–275.

Rees Philip, Sander Nikola, 2020, “The Role of Internal Migration as a Driver of Regional Population Change in Europe,” Comparative Population Studies, 44, pp. 533-544.

Rönnlund Maria, Tollefsen Aina, 2024, “School-to-work transitions in rural North Sweden: staying on in a reviving local labor market,” Journal of Youth Studies, 27(9), pp. 1358–1375.

Ryser Laura, Halseth Greg, Markey Sean, Gunton Cameron, Argent Neil, 2019, “Path dependency or investing in place: Understanding the changing conditions for rural resource regions,” The Extractive Industries and Society, 6(1), pp. 29–40.

Sandow Erika, Lundholm Emma, 2023, “Leaving the City: Counterurbanisation and Internal Return Migration in Sweden,” European Journal of Population, 39(1): 7.

Stein Jonas, 2019, “The Striking Similarities between Northern Norway and Northern Sweden,” Arctic Review on Law and Politics, 10, pp. 79–102.

Strasser Roger, 2016, “Learning in context: Education for remote rural health care,” Rural and Remote Health, 16(2), pp. 1–6.

Sveriges Kommuner och Regioner, 2023, “Classification of Swedish municipalities,” online. Available at: https://skr.se/en/skr/tjanster/kommunerochregioner/faktakommunerochregioner/kommungruppsindelning.2051.html (accessed, 15 April 2025).

Taylor Andrew, Larson Silva, Stoeckl Natalie, Carson Dean, 2011, “The haves and have nots in Australia’s Tropical North - New Perspectives on a Persisting Problem,” Geographical Research, 49(1), pp. 13–22.

Taylor Andrew, Thurmer James, Karácsonyi David, 2022, “Regional demographic and economic challenges for sustaining growth in Northern Australia,” Regional Studies, Regional Science, 9(1), pp. 425–445.

Thurmer James, Carson D., Taylor A., 2019, “Modelling population retention in Australia’s Northern Territory – how do current forms of migration contribute to population turnover and retention?,” Australian Geographer, 50(4), pp. 435–452.

Vuin Ana, 2019, Migration against the tide: Case studies of South Australia, Sweden and Croatia, Doctoral dissertation, Darwin, Charles Darwin University.

Haut de page

Table des illustrations

Titre Figure 1: Sweden’s north SPAs
URL http://journals.openedition.org/eps/docannexe/image/16663/img-1.png
Fichier image/png, 181k
Titre Figure 2: Changes in in-migration intensity, 2002-2023 (baseline 2002)
URL http://journals.openedition.org/eps/docannexe/image/16663/img-2.png
Fichier image/png, 110k
Titre Figure 3: Changes in out-migration intensity, 2002-2023 (baseline 2002)
URL http://journals.openedition.org/eps/docannexe/image/16663/img-3.png
Fichier image/png, 94k
Haut de page

Pour citer cet article

Référence électronique

Dean Bradley Carson et Doris Anna Carson, « Are sparsely populated areas more disconnected from local labor markets? Examining local migration intensities in northern Sweden »Espace populations sociétés [En ligne], 2025/2 | 2025, mis en ligne le 01 décembre 2025, consulté le 17 juin 2026. URL : http://journals.openedition.org/eps/16663 ; DOI : https://doi.org/10.4000/167oj

Haut de page

Auteurs

Dean Bradley Carson

(corresponding author)
Department of Epidemiology and Global Health
Umeå University 
90187 Umeå, Sweden
dean.carson[at]umu.se

Doris Anna Carson

Department of Geography
Umeå University
90187 Umeå, Sweden
doris.carson[at]umu.se

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont susceptibles d’être soumis à des autorisations d’usage spécifiques.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search