Seasonal residence and counterurbanization: the role of second homes in population redistribution in Finland

Seasonal esidence and coun e u baniza ion: he ole o second homes in popula ion edis ibu ion in Finland Czesław Adamiak .Ka i Pi ka ¨nen .Olli Leh onen ÓThe Au ho (s) 2016. This a icle is published wi h open access a Sp inge link.com Abs ac Since he la e wen ie h cen u y, many de eloped coun ies ha e expe ienced popula ion deconcen a ion, labelled as coun e u baniza ion. The e has been an academic discussion on he mean- ing, alidi y and uni e sali y o his concep , d i e s o coun e u baniza ion and i s impac s on u al a eas. To da e, esea ch on coun e u baniza ion mos ly apply s a ic and disc e e de ini ions o esidence, mig a ion and popula ion, which is an inc easingly simplis ic iew in he con empo a y eali y o a g owing mul i- ude o o ms o mobili y, o en ela ed o dual esidence. Pa icula ly la ge-scale quan i a i e s udies on coun e u baniza ion a e con ined by exis ing s a- is ical p ac ices. This pape a emp s o o e come his obs acle and desc ibe he ans o ma ion o he se le- men sys em in Finland acknowledging he ole o second home mobili y. To achie e his goal, i in oduces wo al e na i e measu es o popula ion, seasonal and a e age popula ion, and analyses hei spa ial dynamics be ween he yea s 1990 and 2010 based on geo e e enced g id s a is ical da a. The s udy inds ha al hough egis e ed popula ion has been concen a ing du ing he pe iod in analysis, seasonal popula ion has been inc easingly dispe sed due o he g owing numbe o second homes. I shows ha he coun e u baniza ion p ocess, hough no no iced by con en ional s a is ics, does occu in Finland, mani- es ed by seasonal a he han pe manen mo es. The a icle concludes ha a ious o ms o mobili y should be aken in o accoun when analysing he u ban– u al popula ion dynamics and ans o ma ions o se le- men sys ems as well as in u al de elopmen planning. Keywo ds Second homes Seasonal popula ion  Coun e u baniza ion Popula ion deconcen a ion  Finland In oduc ion Popula ion concen a ion and deconcen a ion a e among he cen al hemes in popula ion geog aphy. Since he beginning o he indus ializa ion e a, u baniza ion has led o an inc easing popula ion concen a ion in u ban a eas, o en pa alleled by he depopula ion o u al egions. Since he 1970s, an opposi e coun e u baniza ion end has been obse ed in many de eloped coun ies, in e p e ed as a esul o C. Adamiak (&) Facul y o Ea h Sciences, Nicolaus Cope nicus Uni e si y, To un ´, Poland e-mail: [email p o ec ed] K. Pi ka ¨nen En i onmen al Policy Cen e, Finnish En i onmen Ins i u e (SYKE), Joensuu, Finland e-mail: [email p o ec ed] O. Leh onen Ru al Policy and Ru al Resea ch, Na u al Resou ces Ins i u e Finland, Helsinki, Finland e-mail: [email p o ec ed] 123 GeoJou nal DOI 10.1007/s10708-016-9727-x he adi ionally dominan economy-d i en u al- o- u ban mo emen s being ou numbe ed by u ban- o- u al mig a ions o ien ed on highe quali y o li e (Be y 1976; Champion 1989). The inc ease in esea ch on coun e u baniza ion has b ough concep- ual con usion a ound he opic mani es ed in an essen ial disc epancy in he unde s anding o coun- e u baniza ion ei he as a o m o indi iduals’ mig a o y mo emen om u ban o u al a eas, o as a p ocess o se lemen sys em ans o ma ion in o a mo e dispe sed o m (Mi chell 2004). S udies on coun e u baniza ion usually adop s a ic de ini ions o esidence (as a ixed cha ac e is ic o a pe son, which can be shi ed in he e en o mig a ion) and popula ion (as a cha ac e is ic o a place, which can change o e ime, bu ideally is measu eable a any momen ). Such assump ions a e being inc easingly challenged by con empo a y social science on human mobili y (Hall 2005;U y2007). I has been sugges ed ha he no ions o home, place o esidence, mig a ion, and popula ion do no su ice o unde s and inc easingly di e se o ms o mobili ies, o en ela ed o ou ism and mul iple esidence (McIn y e e al. 2006). A emp s o acknowledge empo a y mobili y in popula ion mea- su es ha e been mainly made wi hin ou ism s udies (Te ie 2006; Rigall-i-To en 2010). Mu ¨lle and Hall (2003) es ima ed seasonal popula ion edis ibu ion ela ed o second home use in Sweden. Howe e , he pu sui s o desc ibe seasonal popula ion lack connec ion o he schola ly li e a u e on u ban– u al popula ion dynamics. Such linkage could en ich he discussion on coun e u baniza ion which, as Hal ac ee (2012)sug- ges s, should no be con ined o he s udy o pe manen mig a ions only, bu should ins ead conside he whole spec um o empo a y mobili ies om commu ing h ough leisu e isi s and second home use. In he No dic coun ies, s a is ics on egis e ed popula ion show ha he concen a ion in u ban a eas con inues o domina e o e a con a y popula ion low (Heikkila ¨2003; Leh onen and Tykkyla ¨inen 2009). In Finland, in pa icula , he ex ensi e u banisa ion p ocess is o ela i ely ecen o igin, as i s a ed only in he 1960s, and since hen i has caused nega i e impac s on he economic de elopmen p ospec s and li ing en i- onmen s o u al a eas (Leh onen 2015;Leh onenand Tykkyla ¨inen 2010). I has been hoped ha he decline in u al popula ion and income base could be o se by mig a ion om ci ies, especially o e i ed households looking o e u n o hei u al oo s and li es yle mig an s (Jauhiainen 2009). Howe e , no la ge-scale u ban o u al mig a ion has aken place, one o he explana ions being ha he quali y o li e mo i es, elsewhe e linked o u ban- o- u al mig a ion, in he No dic egiona e a he sa is iedby he ex ensi e access o u al second homes (Mu ¨lle 2011; Niedomysl and Amco 2011;Hil unene al.2013). I has been es ima ed ha hal o he popula ion o he No dic coun ies has an access o a second home and hese a e inc easingly used yea - ound. Hence, ins ead o pe ma- nen mo es people inc easingly op o sha ing hei li es be ween an u ban pe manen esidence and a u al second home (Adamiak e al. 2015). Second homes ha e la gely been igno ed in u al policies and local planning based on egis e ed pe manen popula ion igu es, al hough a g owing numbe o s udies ha e poin ed a hei impo ance as a way o balance he nega i e e ec s o depopula ion o u al communi ies and secu e hei u u e by helping o p ese e se ices and employmen (Mu ¨lle 2002;Ma ja aa a2008; Hil unen e al. 2013). This pape explo es he spa ial pa e ns o popula ion de elopmen in Finland be ween 1990 and 2010. We p esen a way o acknowledge he ole o empo a y mobili y in he p ocess o se lemen sys em ans o ma- ion, in esponse o he appeals o including he di e si y o human mobili ies in o he popula ion geog aphy. Mo eo e , we aim o e eal he spa ial pa e ns o popula ion deconcen a ion hidden unde he phe- nomenon o second homes. To achie e hese goals, we in oduce wo al e na i e measu es o popula ion: seasonal and a e age popula ion, and analyse hei dynamics be ween 1990 and 2010 using g id s a is ical da a, complemen ed wi h mo e in-dep h su ey da a on he access and use o second homes in Finland. We use Finland as case s udy o h ee easons: he s ong and ongoing u baniza ion p ocess ha he coun y is going h ough and ela ed weakening de elopmen p ospec s o he u al a eas; he la ge scale o second home mobili y; and he a ailabili y o unique GIS and su ey da a ha enabled us o pe o m a de ailed quan i a i e s udy. Backg ound Coun e u baniza ion as mig a ion and a p ocess o se lemen sys em ans o ma ion U baniza ion, de ined as he p ocess o inc easing concen a ion o popula ion in ci ies (Tisdale 1942) GeoJou nal 123 has ans o med Eu ope om p edominan ly u al in o u ban du ing he ecen wo cen u ies. In many coun ies, massi e mig a ions om u al a eas o ci ies lead o popula ion decline in u al egions unde mining hei economic pe o mance and causing ad e se social consequences ( an de Be g e al. 1982; Dahms 1995; S ockdale 2002; Rye 2006). Mos o he de eloped coun ies eached he ceiling o apid u baniza ion in he cou se o he wen ie h cen u y, and he unidi ec ional ans o ma ion ga e way o mo e complex pa e ns o u ban– u al popula ion dynamics (Champion 2001). In he 1970s, a e e sal o he u ban concen a ion p ocess was no iced in he USA and many wes e n Eu opean coun ies, which many schola s in e p e ed as a sign o a u na ound o he dominan end in o popula ion deconcen a ion. Coun e u baniza ion was in e p e ed o be d i en by he inc easing impo ance o quali y o li e and pos - Fo dis economic ans o ma ion (Be y 1976; Cham- pion 1989; Va iainen 1989a). The p edic ions o deconcen a ion becoming a widesp ead and dominan popula ion endency (Vin- ing and S auss 1977) ha e been la gely c i icized by la e s udies. The gene alizabili y o his de elopmen has no ound suppo om s a is ical sou ces (Cham- pion 2001). Long-wa e economic ends and non- linea e olu iona y models we e used o explain empo a y, a he han pe manen , cha ac e o popu- la ion deconcen a ion ( an de Be g e al. 1982; Be y 1988; Geye and Kon uly 1993). Mo eo e , i has been deba ed o wha ex en s a is ical obse a ions p o e ac ual u al g ow h a he han a e i o ial expansion o ci ies: u ban spillo e o subu baniza ion (Go don 1979; Kon uly and Vogelsang 1988; Amco 2006). Also, he popula ion dynamics be ween u ban and u al a eas a e nowadays a ibu ed o a wide spec um o gene al and place-speci ic ac o s shaping u ban– u al mo emen s, bu also in e na ional mig a ions and na u al inc ease (Champion 1989; Dahms 1995; Hogga 1997; Mi chell 2004; Johnson e al. 2005). The e m coun e u baniza ion is nowadays used no only o name a p ocess o se lemen sys em ans o - ma ion, bu also a mig a o y mo emen om u ban o u al a eas (Mi chell 2004). Such mig a ion may, bu does no need o change he u ban– u al popula ion balance, because i is usually accompanied by an opposi e u al- o-u ban mig a ion, as well as o he mig a o y and na u al mo emen s. Ye , i al e s he s uc u e o u al popula ion and b ings a ious economic and social consequences o u al communi- ies (Hogga 1997; Boyle and Hal ac ee 1998; Milbou ne 2007; Phillips 2010). U ban- o- u al mo e- men has adi ionally been a ibu ed o ela i ely weal hy amilies o middle-aged adul s seeking a mo e na u al and elaxed dwelling en i onmen (Boyle and Hal ac ee 1998; Gosnell and Ab ams 2011;S ˇimon 2012). Locally, such mig a ion is mos commonly di ec ed o a eas o special na u al appeal as sea coas s and moun ains (Magnusson and Tu ne 2003; McG anahan 2008), and hus labelled as ameni y mig a ion (Gosnell and Ab ams 2011; Moss 2006). S ill, con empo a y u ban- o- u al mig a ion is no limi ed o he middle-class quali y o li e seeke s. Conside able g oups a e also d i en by economic ac o s: employmen possibili ies, lowe cos s o li e o li e e en s such as e i emen (including e u n mig a ions), unemploymen , indeb edness, ma iage, di o ce (Mi chell 2004; Pehkonen 2005; S ockdale 2006; Hal ac ee 2008; Jauhiainen 2009;S ˇimon 2012). Also in e na ional immig a ion, which adi ionally ocused on u ban a eas, is inc easingly a ge ed o u al a eas (Milbou ne 2007; Hedbe g and Haand ik- man 2014). The e is an inconsis ency in de ining coun e u - baniza ion, wi h basic disc epancy be ween he indi- idual mobili y app oach and he se lemen sys em app oach. To sol e his p oblem Mi chell (2004) p oposed o apply h ee e ms ins ead o a single one and she buil a concep ual model explaining causal ela ions be ween hem. She de ined coun e u ban- iza ion as he mig a o y mo emen om u ban o u al a eas. I is one o he ac o s, along wi h na u al inc ease and o he domes ic and in e na ional mig a- ions, ha may cause coun e u banizing p ocess, unde s ood as a p ocess o deconcen a ion o se le- men sys em, which e en ually leads o he c ea ion o acoun e u ban se lemen pa e n. Al hough he usabili y o hese e ms may be deba ed, such a dis inc ion is help ul o discussing how second home mobili y can be sui ed wi hin hese concep s. Coun e u baniza ion and second homes Many o he mo i es ha explain mig a ions om u ban o u al a eas, such as escape om u ban li e, he appeal o na u al en i onmen , and willingness o own o build a home, a e also iden i ied as impo an mo i es o he owne ship o second homes (Jaakson GeoJou nal 123 1986; Kal enbo n 1998; Hall and Mu ¨lle 2004). The de elopmen o second homes also aises simila issues o social and economic in e ac ions as pe ma- nen mig a ion and second home use some imes leads o a pe manen mo e in o u al a ea (Williams and Hall 2000; Ma ja aa a and Lundholm 2014). In ac , some p e ious s udies ha e ea ed second home owne ship and seasonal esidence as a mani es a ion o ameni y mig a ion (McCa hy 2008) o coun e u - baniza ion (Bulle and Hogga 1994). Re e ing o he concep ualiza ion o coun e u baniza ion Hal ac ee (2012) has p essed o acknowledging he wide spec um o mobili y om ansi isi s h ough seasonal second home use o pe manen esidence, as all ha e common mo i es and impac s bu di e in hei empo al scope. I hus seems ha esea ch in coun e u baniza ion, unde s ood as mobili y (Mi chell 2004), has mo e o less consciously accep ed second homes as a pa o his phenomenon. P e ious s udies on coun e u baniza ion as a se - lemen sys em ans o ma ion (coun e u banizing) p ocess (Mi chell 2004), in u n, a e mos ly based on la ge-scale egis e -based quan i a i e analyses. Hence, hey ely on s a is ical p ac ices which ea places o esidence as s a ic and disc e e poin s, and popula ion as a measu able cha ac e is ic o a e i o- ial uni . A emp s o measu e popula ion dis ibu ion and dynamics acknowledging empo a y and seasonal mobili y o en come om ou ism s udies (Bell and Wa d 2000; Te ie 2006). Some esea ch employ indi ec me hods o assessing seasonal popula ion changes such as a el su eys (Cha les-Edwa ds and Bell 2015), da a on he usage o wa e , elec ici y and was e p oduc ion (Rigall-i-To en 2010) and mobile phone acking (Silm and Ahas 2010), bu he possibili ies o hei applica ion a e usually limi ed in space and ime. Mu ¨lle and Hall (2003) es ima ed he adjus ed dis ibu ion o popula ion in Swedish municipali ies aking in o accoun he dis ibu ion o second homes, places o pe manen esidence o hei owne s, a e age numbe s o isi o s and a e age yea ly ime spen in second homes. The esul ing popula ion es ima ions we e signi ican ly di e en han he o icial numbe s: u ban municipali ies los , while some pe iphe al u al municipali ies gained o e 10 % mo e esiden s. Following Mi chell’s (2004) concep ualiza ion we can say ha he au ho s e ealed a mo e coun e u ban se lemen pa e n hidden unde he appa en ly mo e concen a ed one sugges ed by he o icial popula ion s a is ics. Ye he au ho s pe - o med a s a ic analysis and did no in es iga e he coun e u banizing p ocess, hus how he se lemen sys em has changed o e ime due o he second home use. The measu emen o seasonal a iabili y o popu- la ion has signi ican p ac ical implica ions o u al policy. In mos coun ies, people can egis e only one o icial place o esidence whe e hei ci izenship igh s a e ied o. These o icial egis e ed popula ion igu es wo k as he basis o ax dis ibu ion, o en pu ing communi ies wi h high numbe s o seconda y dwellings in an un a ou able si ua ion (Mu ¨lle and Hall 2003). The need o ake in o accoun seasonal popula ion in u al planning policies was no iced al eady in he 1970s (Raga z 1970; Coppock 1977) and i is u he augmen ed by hopes and ea s associa ed wi h he de elopmen o second homes. These a e simila o hose ela ed o pe manen mig a ion in o u al a eas. On he one hand, second home use s o some deg ee eplace popula ion mo ing ou , help o main ain social ab ic and demand o local se ices. On he o he hand, he de elopmen o second homes may lead o con lic s wi h local popula ion, and compe i ion o esou ces some imes esul ing in he displacemen o local esiden s. The impac s o second home de elopmen a y depending on na ional and local con ex (Hall and Mu ¨lle 2004; Gallen e al. 2005; Fa s ad 2013; Hall 2014). U ban– u al popula ion dynamics and second homes in Finland In ecen decades, Finnish popula ion has inc easingly concen a ed in ci ies, which has been accompanied by he depopula ion o u al a eas (Ales alo 1983; Kupiszewski e al. 2000; Pekkala 2003). A symp om o popula ion deconcen a ion has been no iced in he 1970s when he popula ions o some pe iphe al egions in cen al and no he n Finland s a ed o inc ease, bu i happened only hanks o he g ow h o main ci ies o hese egions, while he u al a eas we e con inuously loosing popula ion (Va iainen 1989b; Heikkila ¨2003;Ha ¨ a ¨la ¨and Rusanen 2010). U ban- o- u al mig a ion in Finland is a ied in e ms o ime, space, mo i es and social composi ion. I includes e u n mig a ion o e i ed people, olun a y and economic mig a ion o amilies o subu ban a eas and o me esidences o pe iphe al a eas. I does no GeoJou nal 123 howe e su ice o supplemen he domina ing u al- o-u ban low (Heikkila ¨2003; Pekkala 2003; Pehko- nen 2005), esul ing in con inuous depopula ion and economic s agna ion o spa sely popula ed pe iphe al a eas (Leh onen and Tykkyla ¨inen 2010; Leh onen 2015). To desc ibe he dynamics o egis e ed popula ion in Finland since he 1980s, we use he YKR u ban– u al ypology, which di ides Finnish e i o y in o se en ca ego ies based on a complex classi ica ion p ocedu e using da a on popula ion, labou , commu - ing, building and land use o measu e densi y, accessibili y, e sa ili y and o ien a ion o he a eas (Helminen e al. 2014; see Da a and Me hods o he desc ip ion o he YKR geo e e enced s a is ical da abase). F om 1980 o 2012 u ban and exu ban a eas ha e been cha ac e ized by a popula ion inc ease (Fig. 1). The exu ban inge a eas sp ead a ound and be ween u ban cen es and a e linked o hem by in as uc u e and unc ional ela ions. In pa allel o he popula ion g ow h in u ban and exu ban a eas, he ou ca ego ies o u al a eas in YKR ypology ha e been mos ly cha ac e ized by a loss o popula ion. Only u al a eas close o u ban a eas ha e wi nessed a s eady popula ion g ow h du ing he pas decades. Local adminis a i e and se ice cen es in u al a eas ou side o he each o u ban in luence ha e had ela i ely s able popula ion numbe s. The emaining wo u al ca ego ies: co e u al a eas and spa sely popula ed u al a eas, which oge he co e 84.2 % o he Finnish land a ea, ha e expe ienced a signi ican loss o popula ion. Second homes cons i u e an impo an pa o housing and land use in Finland. They a e mos ly loca ed in u al a eas, and, in con as o pe manen esidencies, hei numbe in hese a eas has been g owing du ing he pas decades. Acco ding o S a is- ics Finland, since 1970 he numbe o second homes has almos ipled om 176 o 500 housand in 2014, and e e y i h o he 317 Finnish municipali ies has mo e second homes han pe manen dwellings (OSF 2015). The impo ance o second homes is u he emphasized by he ac ha second homes a e no only used by hei owne s, bu also by ex ended amily and iends. Acco ding o he Longi udinal Na ional Ou doo Rec ea ion Su ey (LVVI 2010), abou 40 pe cen o Finns ha e a egula access o a second home, hus i may be es ima ed ha each second home in Finland is egula ly used by 4.2 people. I also andom isi s a e coun ed, al oge he o e 3 million Finns may isi a second home annually (Nieminen 2009). The access a e is sligh ly highe among u ban popula ion and u ban esiden s spend mo e ime in second homes (Table 1). On a e age, hose wi h an access o a second home spend he e 43 days a yea . Acco ding o he LVVI s udy, bo h he sha e o Fig. 1 Change o egis e ed popula ion in Finland be ween 1980 and 2012 by ca ego y o YKR u ban– u al ypology GeoJou nal 123 popula ion wi h egula access o second homes and a e age ime spen he e ha e been g owing du ing he pas decade. Da a and me hods Popula ion s a is ics on sub-na ional le el a e usually ga he ed o e i o ies o adminis a i e uni s (e.g. municipali ies). An al e na i e app oach is o di ide he e i o y in o a egula g id o small ec angles o he same size. Popula ion igu es and o he da a a e assigned o ce ain ec angles (g id cells) based on he exac geo e e enced loca ions o places o esidence, p ope ies o companies. S a is ical g id da abases a e inc easingly used in popula ion s udies in he No dic coun ies (Amco 2006; Ko a aa a e al. 2011). Pe o ming spa ial popula ion analysis independen ly om adminis a i e bo de s helps o a oid se e al me hodological pi alls, such as he e ec o ‘‘u ban spillo e ’’ (Amco 2006) o la ge adminis a i e uni s co e ing dispa a e a eas: in Finland, especially in he no he n pa o he coun y, municipali ies co e la ge a eas which may be p edominan ly dese bu con ain municipal owns whe e local popula ion is concen- a ed (Muilu and Rusanen 2003; Kauppila and Rusanen 2009). In ou analysis, we use he YKR da abase (Mon- i o ing Sys em o Spa ial S uc u e and U ban Fo m), one o he mos ad anced geo e e enced s a is ical da abases in he wo ld enabling o pe o m na ionwide analyses using GIS me hods. The YKR da abase is c ea ed and main ained by he Finnish En i onmen Ins i u e, and i comp ises da a abou popula ion, housing, wo kplaces and a els o wo k om he yea s 1980 o 2010 o each o abou 6.3 million 250 m 9250 m squa e-shaped g id cells in o which he e i o y o Finland is di ided. We gene alized he o iginal da a in o lowe spa ial esolu ion o 5km95 km cell size, as such esolu ion is su icien o he pu pose o he s udy and gene aliza ion acili a ed compu ing and g aphical p esen a ion o he esul s. F om he YKR da abase, we de i ed he in o ma- ion on he numbe s o egis e ed popula ion and p ope ies desc ibed as second homes in each g id cell. The e m second home e e s o buildings used as a leisu e- ime esidencies on he las day o he yea (OSF 2015). Due o sligh di e ences in s a is ical me hodologies, he YKR da a di e s om he da a by S a is ics Finland. The o al numbe o second homes in he YKR da abase in 2010 was 517,622. This numbe con ained 489,232 (94.5 %) indi idually used second homes and 28,390 (5.5 %) comme cial second homes en ed ou o ou is s. The comme cial second homes could no be excluded om he analysis, because due o p i acy p o ec ion he da abase did no p o ide he de ailed p opo ions be ween he wo ca ego ies in pa o g id cells. On he o he hand, he da abase does no include all p ope ies used as second homes, as a ms con e ed in o second homes, la s used as second homes, allo men ga den co ages and en al second homes migh be egis e ed o o he pu poses, usually as esiden ial buildings (OSF 2015). Also, he ans o ma ions o he use o p ope ies be ween p ima y and seconda y esidences a e o en no egis e ed. In ou analysis we compa e he dis ibu ion o popula ion in 1990 and 2010. The choice o ha ime span enables us o obse e he dynamics in he popula ion dis ibu ion du ing he pe iod o economic g ow h ha is said o ha e con ibu ed o he accele a ed concen a ion o Table 1 Ra e o access and a e age numbe o days spen a second home by size o municipali y o esidence. Sou ce: LVVI (2010) Size o municipali y (inhabi an s) Ra e o access o second homes (% o popula ion) A e age numbe o days a yea spen a second home 4000 37.1 37.2 4000–9999 39.8 37.7 10,000–24,999 38.8 41.4 25,000–99,999 41.7 44.7 C100,000 41.9 45.9 Na ional a e age 40.8 43.3 GeoJou nal 123 popula ion in majo u ban a eas (Leh onen and Tykkyla ¨inen 2009). In o de o ake in o accoun he ole o second home use in he popula ion dis ibu ion, we analyse no only he changes in he egis e ed popula ion numbe s (RP), bu we also elabo a e wo addi ional measu es o popula ion. The seasonal popula ion (SP) measu es he numbe o people expec ed o be p esen in gi en a ea du ing he highes ou is season. The a e age popula ion (AP), in u n, measu es he a e age numbe o people p esen a he a ea h oughou a yea . To calcula e hese measu es, we use wo pa ame e s: he a e o access o second homes and he a e age leng h o s ays in second homes, es ima ed o esiden s o di e en a eas in Finland (Table 1), and based on he esul s o he Longi udinal Na ional Ou doo Rec ea ion Su ey (LVVI 2010) conduc ed in he yea s 2009–2010 and a ge ed o a andom sample o almos 9 housand Finns. To assign hese pa am- e e s, we di e en ia ed g id cells acco ding o he size o municipali ies hey belonged o. Be o e calcula ing seasonal and a e age popula- ion, we assigned numbe s o second home use s (SHU) o each g id cell. Fo he yea 2010 i was calcula ed as SHU2010i¼sh2010iuse s and o he yea 1990 as SHU1990i¼sh1990iuse s In hese equa ions SHU2010 i and SHU1990 i a iables ep esen numbe s o he use s o second homes loca ed in a gi en g id cell iin he yea s 2010 and 1990 espec i ely. The sh2010 i and sh1990 i a iables s and o he numbe s o second homes in cell iin 2010 and 1990. Fo 2010 we used he eal numbe o second homes acco ding o YKR da a, and o 1990 we used he numbe o second homes exis ing in 2010, which we e buil un il 1990 (hence mino inaccu acies may esul om possible demoli ion, abandonmen o change o use o second homes be ween hese da es). The cons an use s deno es he a e age numbe o people using one second home, which is 4.2, based on he LVVI s udy esul s. I is assumed o be cons an o e ime due o lack o compa able da a om 1990. Seasonal popula ion (SP) coun s he popula ion assuming ha e e yone who has access o a second home is p esen a his/he second home, and no in he place o his/he pe manen esidence. I is no an ac ual numbe o popula ion p esen a he a ea in any momen o ime, because no e e yone isi s second home a he same ime, bu conside ing he high uni o mi y o he seasonal pa e ns o second home use in Finland (summe use o second homes p edomi- na es in all egions o Finland, including he No h whe e many second homes a e loca ed in ski eso s, see Adamiak e al. 2015), i is a ai app oxima ion o popula ion dis ibu ion a he peak o he summe season (in July). Seasonal popula ion is calcula ed by adding he numbe o second home use s o egis e ed popula ion, and sub ac ing he numbe o popula ion in he cell expec ed o lea e hei pe manen esidence o isi a second home. Fo he yea 2010: SP2010i¼RP2010iþSHU2010i RP2010iaccess2010i and o he yea 1990: SP1990i¼RP1990iþSHU1990i RP1990iaccess1990i In hese equa ions SP2010 i and SP1990 i a iables ep esen seasonal popula ion o a gi en g id cell iin he yea s 2010 and 1990 espec i ely. RP2010 i and RP1990 i desc ibe egis e ed popula ion in a simila way. The a iables access2010 i and access1990 i deno e he sha es o g id cell popula ion ha had access o a second home in 2010 and 1990. Fo 2010, hese igu es a ied ac oss municipali ies o di e en sizes based on he LVVI s udy (see Table 1). Fo 1990, hey we e es ima ed based on he numbe o second homes in 1990, wi h he assump ion ha he a e age numbe o use s o one second home, as well as he ela i e di e ences in access o second homes be ween municipali ies o di e en sizes did no change un il 2010. The a e age popula ion (AP) desc ibes he popula- ion a e aking in o accoun he annual use pa e ns o p ima y and second homes. Fo 2010 i was calcula ed as AP2010i¼RP2010iþSHU2010idays 365 RP2010iaccess2010idaysi 365 and o he yea 1990 as GeoJou nal 123 AP1990i¼RP1990iþSHU1990idays 365 RP1990iaccess1990idaysi 365 The cons an days desc ibes he na ional a e age o he numbe o days spen in second homes by hei use s annually (43), and o calcula ing he seasonal ou low o popula ion his pa ame e was assigned o a gi en g id cell (days i ) acco ding o he di e ences be ween municipali ies o di e en size (Table 1). We used he same alues o days o bo h yea s: 1990 and 2010 due o he lack o p ecise da a on he a e age ime pa e ns o second homes use in 1990. Howe e , he mean numbe o days spen in second homes inc eased om 31 o 43 only be ween he yea s 2000 and 2010 (LVVI 2000,2010), so ou es ima ion o he change in a e age popula ion be ween 1990 and 2010 may be oo low. Change in he egis e ed popula ion be ween 1990 and 2010 in s a is ical g id cell i(DRP i ) was calcula ed as DRPi¼RP2010iRP1990i and changes in seasonal and a e age popula ion we e calcula ed acco dingly. We compu ed he change o each measu emen o popula ion o each ca ego y o YKR u ban– u al ypology. Gene alized g id cells we e classi ied in o YKR u ban– u al classi ica ion based on he maxi- mum a ea o a single classi ica ion ca ego y inside he cell. The excep ion was made o he local cen es o u al a eas: a g id cell was classi ied as such i o e 10 % o i s a ea ell in o his class. To p esen he changes in he o e all le el o concen a ion o each measu emen o popula ion, we used he Hoo e concen a ion index o en applied in popula ion s udies be o e (Hoo e 1941; Roge son and Plane 2012). Hoo e index (H) is calcula ed as hal o he sum o di e ences be ween sha es o popula ion and a eas o each e i o ial uni (p i and a i espec i ely) in o al popula ion and a ea o he coun y (Pand A): H¼1 2X n i¼1 pi Pai A      Hoo e index alue equals 0 when he whole popu- la ion is dis ibu ed uni o mly in all e i o ial uni s and app oaches 1 (100 %) when he whole popula ion is concen a ed in a small a ea. The index alues can be in e p e ed as a sha e o popula ion ha needs o be eloca ed in o de o ob ain i s uni o m dis ibu ion. Resul s The spa ial pa e ns o h ee measu es o popula ion desc ibed abo e di e signi ican ly be ween each o he . Regis e ed popula ion (Fig. 2, le map) is concen a ed in la ge u ban a eas, mos ly loca ed in sou he n Finland (Helsinki, Tampe e, Tu ku) and egional ci ies in cen al and no he n pa s o he coun y (e.g. Jy a ¨skyla ¨, Kuopio, Oulu), while he es o Finland is less o spa sely popula ed. The dis ibu- ion o second homes and ela ed addi ional popula ion ha can isi he a ea in summe season is di e en (Fig. 2, igh map): i is concen a ed on he coas and a chipelagos on he Bal ic Sea, in he Lakeland egion co e ing he sou he n and sou h-eas e n in e io o he coun y, and in some ou ism cen es in no he n Finland (e.g. Kuusamo, Le i and Vuoka i). I can be no iced ha ou side o u ban cen es, in a signi ican pa o he coun y second home use s may ou numbe pe manen esiden s du ing high season. The spa ial pa e n o seasonal popula ion is a ec ed bo h by he dis ibu ion o pe manen popu- la ion, and second homes (Fig. 3, le map). Seasonal popula ion is concen a ed in u ban a eas, on he Bal ic coas , and in he Lakeland egion. The di e ence be ween pe manen and o seasonal popula ion is pa icula ly e iden in he case o he Lakeland egion which is a popula second home a ea, easily accessible om he Helsinki egion. Taking seasonal popula ion in o accoun also dec eases he ex en o unpopula ed a ea in Finland om almos 1/3 o i s e i o y, sugges ed by he egis e ed popula ion igu es, o less han 1/4 (Table 2). The spa ial pa e n o he a e age popula ion (Fig. 3, igh map) is simila o he pa e n o egis e ed popula ion. Di e ences may be no iced in a eas wi h la ge concen a ions o second homes, such as he coas line, eso s in no he n Finland, and some pa s o he Lakeland, whe e popula ion coun is signi ican ly highe i we conside seasonal in lux. Be ween 1990 and 2010, he egis e ed popula ion o Finland g ew by 7.7 % om 4935 o 5317 housand. The g ow h was spa ially concen a ed: only 16.1 % o he Finnish e i o y expe ienced an inc ease in egis e ed popula ion, while popula ion dec eased on GeoJou nal 123 almos hal o he a ea o he coun y (Table 2). In con as , seasonal popula ion inc eased in he majo i y o he Finnish e i o y. The a ea whe e he a e age yea ly popula ion g ew was geog aphically o e wice as big as he a ea whe e he egis e ed popula ion g ew, and nea ly equalled he a ea whe e a e age popula ion dec eased. The di e ence in he inc ease in egis e ed and seasonal popula ion is a esul o he g ow h o he numbe o second homes by 36.2 % in he pe iod in ques ion: om 380 o 518 housand (acco ding o he YKR da a). The numbe o people ha seasonally change hei place o esidence can be hus es ima ed o ha e isen om 1.59 o 2.17 million be ween 1990 and 2010. The geog aphic pa e ns o changes in he h ee measu es o popula ion be ween 1990 and 2010 a e p esen ed in Fig. 4. In he case o he egis e ed popula ion, he g ow h was mos ly limi ed o u ban and subu ban a eas o Helsinki, Tampe e, Tu ku, Oulu, Jy a ¨skyla ¨, and o he egional cen es. The es o he coun y expe ienced depopula ion. Seasonal popula ion, in u n, inc eased also in he u al a eas o he Lakeland, coas s, and in la ge a eas in he No h o he coun y. Seasonal popula ion dec eased only in he ag icul u al wes e n Finland and in a eas close o he eas e n bo de . The spa ial pa e n o he changes o a e age popula ion is simila o he changes o he egis e ed popula ion, bu he depopula ing a eas a e smalle since he depopula ion in he ameni y- ich a eas is compensa ed by he inc easing seasonal popula ion. The esul s indica e ha hese a eas ha e no been as depopula ed as he s a is ics o egis e ed popula ion sugges . Supe imposing he geog aphic pa e ns o he dynamics o di e en popula ion measu es on he YKR u ban– u al classi ica ion con i ms ha he popula ion de elopmen be ween 1990 and 2010 has been ela ed o he se lemen hie a chy (Table 3). Fig. 2 Dis ibu ion o egis e ed popula ion and second home use s in Finland in 2010 GeoJou nal 123 Va iainen, P. (1989b). The end o d as ic depopula ion in u al Finland: E idence o coun e u banisa ion? Jou nal o Ru al S udies, 5(2), 123–136. Vining, D. R. J., & S auss, A. (1977). A demons a ion ha he cu en deconcen a ion o popula ion in he Uni ed S a es is a clean b eak wi h he pas . En i onmen and Planning A, 9(7), 751–758. doi:10.1068/a090751. Williams, A. M., & Hall, C. M. (2000). Tou ism and mig a ion: New ela ionships be ween p oduc ion and consump ion. Tou ism Geog aphies, 2(1), 5–27. doi:10.1080/146166800363420. GeoJou nal 123