Cross-border commuting along the Dutch border: Regional and industrial variation in incoming commuting from Germany and Belgium Arjen Edzes, Viktor Venhorst & Jouke van Dijk University of Groningen, Department of Economic Geography, PO Box 800, 9700 AV Groningen Lisbon, ERSA, 27-08-2014
2 Motivation 1. Cross-border issues (economy, safety, labour market, migration, governance) are a big theme in Europe (INTERREG) 2. Cross-border commuting can stimulate cross-border regional economies and solve discrepancies on the cross-border labour market 3. Specifically for the Dutch: cross-border issues of population decline and local financial responsibility for unemployed (exporting unemployed) 4. Lack of insight in actual figures but also lack of insight in drivers and impact of crossborder commuting 5. In this paper we will look at incoming commuters from Germany and Belgium and the relation with industrial and regional characteristics
Population density 2012 Urban-rural typology 2012 4
Cross-border commuting 2012 Source: European Labour Force Survey. NARSC November 2010 5
Relevancy, theory and literature / different perspectives Individual perspectives versus regional perspectives: What moves people and what do they gain? Potential and Impact for regions Neo-classical economic perspectives versus sociological/socialconstructivist perspectives. What dominates the pattern: Economics (wages, job opportunities) or Culture (familiarity, language) 6 Complex interactions between housing and labour markets Shares of incoming commuters, 2012 100%# 90%# 80%# 70%# 60%# Rest# 50%# Dutch#naAonality# German#naAonality# 40%# Belgian#naAonality# 30%# 20%# 10%# 0%# Belgium# Germany# Total# Total of incoming commuters, 2008-2012 100.000" 90.000" 80.000" 70.000" 60.000" 50.000" 40.000" 30.000" 20.000" 10.000" 0" 2008" 2009" 2010" 2011" 2012" total" BE" DE" NL" OV"
Exploratory research questions: 7 1. Do Dutch industrial sectors lean on incoming commuters from Belgium or Germany? 2. Do regional (labor market) characteristics explain the level of incoming commuters?
Expectations Catchment-area: incoming commuting more in border-regions Industrial sectors: variation in dependency of incoming commuters (no clear theory) Labour market: positive relations with job opportunities Housing-prices: relations with mean housing value (for Dutch positive, for Belgian and German negative) More western migrants have positive relations with incoming commuting More productive and smart regions attracts incoming commuters 8
The Netherlands 6.907.400.. employees Germany 27.550.500 employees Belgium 3.836.800 employees Weterings, A. & G. van Gessel-Dabekaussen (2015) Arbeidsmarkt zonder grenzen. PBL/CBS
10 Cross-border commuters with Belgium nationality Cross-border commuters with German nationality Weterings, A. & G. van Gessel-Dabekaussen (2015) Arbeidsmarkt zonder grenzen. PBL/CBS
Data 11 1. An aggregated dataset over 2008-2012 of persons who work in the Netherlands but live in Belgium or Germany (Nationality, Industry - Nace-2 digit, Region of work (NUTS-3) 2. Shortcomings: Only employed workers, based on tax registrations. No information on workers who are temporarily posted. 3. Addition of data on NUTS-3 level: employment * sector, mean housing value, GDP per capita, occupational levels, unemployment rate, proportion of (non) western foreigners 4. All information is coming from Statistics Netherlands.
Estimation approach 12 1. Basic model with only year dummies and dummies for industries: Is there a difference in dependency of incoming commuting between industrial sectors? 2. Addition of regional variables: Do specific regional characteristics lead to more incoming commuting? 3. Analyses of residues: Do regions over- or underperform with regard to incoming commuting? All models estimated for - Commuters from Belgium (1 with Belgium nationality or 2 Dutch nationality) - Commuters from German (3 with German nationality or 4 Dutch nationality)
Descriptives Obs Mean Std, Min Max Commu?ng rcomm_bedetot 3633 0,011 0,027 0,000 0,493 Variables rcomm_be 3633 0,004 0,011 0,000 0,145 rcomm_be with Belgian Nat 3633 0,001 0,005 0,000 0,120 rcomm_be with Dutch Nat 3633 0,003 0,006 0,000 0,057 rcomm_de 3633 0,007 0,024 0,000 0,479 rcomm_de with German Nat 3633 0,003 0,017 0,000 0,435 rcomm_de with Dutch Nat 3633 0,002 0,008 0,000 0,317 Industrial Number of jobs in region*sector / 100000 3633 0,111 0,162 0,000 1,281 Sector var Regional GDP per capita / 100000 3633 0,331 0,088 0,175 0,686 Variables Unemployment rate / 100 3633 0,047 0,009 0,031 0,077 Prop self- employed 3633 0,149 0,016 0,122 0,197 Prop occup at level low 3633 0,085 0,010 0,057 0,114 Prop occup at level medium 3633 0,467 0,042 0,377 0,551 Prop occup at level high 3633 0,180 0,006 0,161 0,200 Prop occup at level scien?fic 3633 0,254 0,053 0,143 0,374 Prop occup at level unknown 3633 0,015 0,010 0,000 0,061 Mean housing value / 1000000 3633 0,230 0,040 0,150 0,357 Prop non- western foreign employees 3633 0,085 0,057 0,025 0,266 Prop western foreign employees 3633 0,085 0,031 0,044 0,189 Border classifica?on Three classifica?on of regions as border regions, non- border but nearby regions and non- border regions: 1) In general, 2) only for the Belgian commuters, 3) Only for the Dutch commuters 13
Belgium Be Nat Belgium NL Nat German Ger Nat German NL Nat Year (ref = 2008) 2009 0,000-0,000*** - 0,001* 0,000 2010 0,000-0,000*** - 0,002*** 0,000 2011 0,000-0,000*** - 0,002*** 0,000 2012 0,000-0,001*** - 0,002*** 0,000 1 digit sector (ref = Agriculture) Mining, quarrying and man 0,002 0,001 0,002** 0,002*** Electricity, gas, steam etc. - 0,001 0,001 0,000 0,006* Water supply 0,000 0,001 0,004 0,003** Construc?on - 0,001 0,001 0,001 0,001** Wholesale and retail trade 0,000 0,001 0,001 0,001** Transporta?on and storage 0,000 0,001 0,004*** 0,002*** Accomoda?on and food service ac?vi??es - 0,001 0,000 0,000 0,000 Informa?on and communica?on 0,000 0,001 0,000 0,001** Financial and insurance ac?vi?es 0,000 0,005** 0,001 0,002*** Real estate ac?vi?es - 0,001 0,002 0,000 0,001 Professional, scien?fic and technical 0,000 0,001 0,000 0,001** Administra?ve and support service 0,001 0,003* 0,033*** 0,006*** Public administra?on and defence - 0,001 0,001-0,001** 0,001** Edua?on 0,000 0,001-0,001 0,001* Human health and social work 0,001 0,001-0,001 0,001 Arts, entertainment and recrea?on 0,000 0,000 0,000 0,001 Other service ac?vi?es - 0,001 0,000-0,001 0,001 Rest - 0,002* - 0,002*** - 0,001** - 0,001*** cons 0,002* 0,002*** 0,002*** 0,001** 14 N 3633 3633 3633 3633 r2 0,028 0,032 0,198 0,041 r2_a 0,022 0,026 0,193 0,035
WITH SECTOR DUMMIES Belgium Be Nat Belgium NL Nat German Ger Nat German NL Nat Year (ref = 2008) 2009 0,000-0,001*** - 0,001* - 0,001*** 2010 0,000-0,002*** 0,000-0,002** 2011 0,000-0,002*** - 0,001-0,001*** 2012 0,000-0,003*** - 0,001-0,003*** Number of jobs in region*sector / 100000 0,001 0,001-0,001 0,002 Region (ref = Border) Non- border, but nearby - 0,003*** - 0,009*** - 0,003-0,005*** Non- border - 0,003*** - 0,011*** - 0,008*** - 0,007*** GDP per capita / 100000 0,004** - 0,002-0,005-0,005* Unemployment rate / 100-0,050 0,130*** - 0,020 0,139** Prop self- employed 0,016*** - 0,011* - 0,048* - 0,005 Prop occup at level low 0,036*** 0,037*** - 0,065-0,064 Prop occup at level high 0,024** - 0,004-0,159** - 0,005 Prop occup at level scien?fic - 0,007** 0,005-0,024-0,013* Mean housing value / 1000000-0,013*** 0,012** 0,002-0,002 Prop non- western foreign - 0,011*** - 0,013*** 0,009-0,005 Prop western foreign 0,069*** 0,044*** 0,042*** 0,005 cons - 0,005* - 0,001 0,052** 0,013 15 N 3633,000 3633,000 3633 3633 r2 0,479 0,645 0,248 0,196 r2_a 0,474 0,641 0,241 0,188
Analysis of residues total incoming commuting
Analysis of residues Belgium
Analysis of residues Germany
19 Conclusion 1. Border regions are most important, but also non-border regions attrackt incoming commuters 2. No clear pattern in industrial sector dependency of incoming commuters from Belgium, for Germany administrative services is important because of Temporarily Employment- and Posting Agencies. 3. No clear effect of labour market characteristics on incoming commuting (not for employment, but preliminary results show also no relation with vacancies). Level of unemployment is only relevant for Dutch cross-border commuters. 4. No effect of total number of jobs 5. Urban-non urban dimensions in regional characteristics; No effect of Smart regions
Thank you for your attention 20