Download fulltext
223

CHILD POVERTY RISKS IN BELGIUM WALLONIA AND FLANDERS: ACCOUNTING FOR A WORRYING PERFORMANCE BY FRANK VANDENBROUCKE* and JULIE VINCK** * professor at the KU Leuven University of Antwerp ** researcher Herman Deleeck Centre for Social Policy 1 INTRODUCTION The at-risk-of-poverty rate children is a ‘lead indicator’ future social problems high child poverty may signal inadequate protection and/or poorly functioning labour markets which be related to lacunae in childcare education system In turn makes success policy more difficult obtain given strong link between economic cultural status their school other words cause effect vicious circle underperforming systems same vein there interplay failing health care Belgium mediocre performer with regard notwithstanding its long tradition security Moreover increasing As first step understanding why our performance worrisome we apply an analytical technique that essence accounting device: decomposition Although this mechanical all simplicity it highlights features Belgian welfare edifice are quite exceptional cross-country comparison but have not been subject much research One these skewed distribution jobs over households Elsewhere study phenomenon by means ‘polarisation analysis’ (Corluy Vandenbroucke 2013a 2013b) paper simple indicator ‘relative severity work poverty’ among Another feature state very work-poor i e little or no participation market These observations show characterised dual polarisation: many live work-poor; simultaneously financial risks 52 BELGISCH TIJDSCHRIFT VOOR SOCIALE ZEKERHEID – 1e TRIMESTER 2015 Section 2 sets scene providing elementary information on Wallonia Flanders 3 illustrates crucially determined ‘work intensity’ household It explains meaning ‘household introduces concepts ‘severe 4 structure intensity poor belong 5 formal analyses; can skipped readers knowing mainly interested results 6 decomposes differences basis using as benchmark country 7 digs deeper into patterns intensity: investigate whether pattern employment explained relatively large share lone-parent 8 another determinant poverty: spending We conclude challenge improve level whilst avoiding ‘(severe) traps’ level: should trigger reconsideration policies several domains 9 decompose increase (as observe since 2005) again do before because developed last decade confronted (disappointing) change meanwhile concluding section sketch three key challenges implying both investment reconsidered light ITS REGIONS use ‘child shortcut defined Eurostat individuals below age 18 Being risk living equivalised net disposable income 60% national median consider one parameters assessment notion ‘poverty’ so used caution reasons This concept presupposes sharing resources within necessarily case applied here rather crude headcount: simply measures threshold does account depth faced ‘at of’ without meaning: present measure signals cut off mainstream society lack headcount defines relation where individual happens living: relative 53 floating changes every year number countries decreased during crisis years reflecting decrease incomes: has favourable impact although needs increased families non-poor alike Our data based EU Statistics Income Living Conditions (EUSILC) ‘2006’ ‘2011’ etc refer SILC survey years; except United Kingdom Ireland they reflect incomes Hence essentially relating 2005 2010 calculations two regions situation particularly alarming Brussels region (and included Belgium) include separate regional analyses EU-SILC sample too small Unless otherwise indicated always refers ‘households’ calculate (based income) thresholds To sure integrated tax benefit only correct relying from normative perspective However calculating rates yields interesting additional intra-regional comparable implies sobering (which less well terms than might assume purely Belgian-wide threshold) likewise harbours inequality rich assume) On 2011 7% quarter Wallonia’s (24 9%) compared 10 4% Applying headcounts 20 8% 13 1% Flanders: Walloon conceived ‘diminishes’ though remains high; Flemish figure hand increases (1) et al (2013 pp 8-10) discuss anchored time For thorough discussion measurement issues involved see Decancq (2014) 54 Figure when European states EUSILC When limit EU15 appears equal unweighted average (18 5%) slightly weighted (19 takes size Comparing nation would constitute category mismatch if interpreted due caution: matter fact also conceal important disparities (Germany telling case) outcome reflects realities different position league German league) FIGURE 1: AT-RISK-OF-POVERTY RATES CHILDREN EUROPEAN WELFARE STATES FLANDERS Notes: ‘FL-FL’ ‘WA-WA’ calculated ‘FL-BE’ ‘WA-BE’ *: significantly 0 05 significance making independent samples t-test Source: compiled authors (2) See (2013) 84-85 illustration argument Germany 55 Given Belgium’s history pioneer record disappointing lead state: just worrying HOUSEHOLD WORK INTENSITY starting point analysis latter measured labelled ratio total months worked working-age members could theory intensities bracket 18-59 considered ‘working age’ students 24 who excluded calculation persons reported having part-time estimate full-time equivalents computed hours habitually interview correlate negatively displays five subsets households: (work ranges 85% 100%) (between 55% 85%) medium 45% 55%) low 20% 45%) (20% less) figures (FL-BE WA-BE) (FL-FL WA-WA) 56 2: THE BELGIAN (BELGIAN REGIONAL LINE) was 8%; 68 ten times higher profile belongs crucial factor explaining his her shows deviates registered 2011: 81 0% subgroup significantly3 average; contrast 3% lower intensity; subgroups difference significant gap haves not’s – ‘have’ referring marginal attachment Rather surprisingly holds Children (3) test Since conservative (4) observation true earlier vintages instance 2008 were 57 face exceptionally reference So highly polarised even ‘internal’ view nuanced: statistically significant; What really particular understood size: working four days week whose 80% couple partner Reconciling family responsibility entail costs two-parents And obviously euros probably former emerges compare distinguishing types broad classes 3: LONE-PARENT HOUSEHOLDS OTHER WITH (BELGIUM EU15) 58 Lone-parents 74 5% 66 Other low: 85 72 short dire straits lone-parents low-medium interval positions change: class (Belgian group 51 0%; 22 nearly disappears (2 6%); loneparent considerably (12 7%) picture similar better question: households? applying definitions 43 14 children; elsewhere (though smaller Is result levels Or above intensity? Across uniform answer some parents Appendix (using explain below): admittedly That justifies focus context But eliminate effort identical metric’ constellations circumstances conclusions return raises fairness adequacy own right 59 now shift attention what extent state’s distinguish indicators (not 20%); label ‘very work-poor’ second (i less); ‘work-poor’ description will third workpoor households; call ‘the introduce must confused poor’ An he/she (financially) poor; thereby ‘working’ minimal (e g being period limited) mixes (is employed employed?) (what income?) intrinsically often leading unwarranted unrelated cases individually counted limited belong; consequence irregularly (Marx Nolan 2014) Diris Verbist examine regression models non-elderly (testing explanatory power severe poverty) covered 2011) turns out combining 60 variables best fit 5These suggests country-specific population concentration play role next general across develops question indicators: comparing develop specific explanations AS AN OUTLIER? subdivide highlight distinction notions workpoor’ ‘income-poor’ legend lowest parts bars represent those combine expressed percentage middle upper work-rich From say adults truly household’s potential valorised nevertheless sum corresponds (5) stress conclusion pooled series whole hold ‘naïve’ presented dependent variable (i) (ii) (iii) transfers pensions bivariate correlation coefficient -0 15 01 deep (in period) (6) provides such indicates (with respect (Vandenbroucke 2013 p 29) 61 4: SUBDIVIDED ON BASIS OF Note: Total differ missing values certain internal Except Hungary Malta belonging driven 6% proportion Wallonia; prompts salient summarises striking (on horizontal axis) vertical Dividing axis 62 5: SEVERE ACROSS disparity poverty’: around Scandinavian 40 Latvia UK Estonia Greece Spain Italy larger Luxembourg Cyprus Poland Switzerland Romania Slovenia Norway; 14% Bulgaria influenced prevalence correlates strongly (negatively) Severe correlated display (7) nonemployment ILO Labour Force Survey 2007 2009 80 90; 50 (exceptionally year) 65 63 diversity patterns; trend line Relative 15% 35% Lithuania 36% 42% Denmark poverty; top range specificity pursue digging Obviously income-poor (when corresponding above) attributable factors: comparatively financially (very) structural disentangle (severe) done shown DECOMPOSITION ANALYSIS: FORMAL DESCRIPTION decompositions characteristic Ptotal written If T k s represents Pk value P each follows: (8) including (2014 18) 64 characteristics under review at-riskof-poverty Sections 9) 7) Equation decomposing intertemporal ∆x (or region) 2006 Formally B indicating Belgium; With follows equation ∆ then term side accounted hypothesis characterising subgroups) identical) ‘differences between’ within’ country) ‘changes (due divided (T=2) reduced following formula example partition (the ‘work-rich’) contributory factors:9 contribution ; ii iii ‘difference comes decomposition; factors together underlying components exercise (taking sizes EU-SILC) admit robust proceeding empirical illustrations caveats taken board Decomposition reveal causality: basically device dissociated presuppose associated definition ‘mechanical’ character overall determinants differences) benchmarking application selection Decom- (9) sections residual linked did available graphs avoid any misunderstanding reason changes) ‘sum decomposition’ residual) observed (10) Using instead yield analysis: explores correlations (examining taking variables); focuses plays most expression ‘cross-country differences’ follow reader read Belgium’: descriptions fortiori combined insights DECOMPOSING CROSS-COUNTRY DIFFERENCES (‘Cross-country’ ) subdivided depending 45 per cent cut-offs ‘Differences 67 6: SUBGROUPS Countries left Norwegian points § Norway ‘between difference’); groups ‘within (0 points) 11 (but Iceland Finland) outperforms various degrees main stems (11) footnote focusing Table A1 7: (TWO SUBGROUPS) Taking (9 decomposed significant) (-0 Finland outperform run Nordic 69 largely Work Spain: seen entirely segment work-poor: estimates confidence intervals them Nevertheless tentative drawn: perform ‘trade-off’ least comparative assessing ex post outcomes macro worse notably Southern (although 12 leads further questions First Wallonia? Second (whilst work-richer segments perspective) explore touch upon (12) equals 19 Wallonia: experience 75 87 line) variant decreases 76 70 wonder chose (by subdividing children) note (see Europe within/between 3) dominates differences’; ‘not emerge go sense respects choice drives remain PATTERNS EMPLOYMENT SIZE STRUCTURE: FURTHER showed expect differences: ‘extended families’ imply degree ‘pooling’ non-employment hence sufficient explanation poverty? dividing 8a reiterates 8b presents (13) income-poverty ‘within’ diminishes Norway/Denmark cf while reverse shape 71 8A: 8B: Figures illustrate observations: marginally (across 36 34 category); 28 corollary children: big majority review); minority 17 2%); iv diversified ceteris paribus how ‘lone-parent’ employment? (including Wallonia) 4) (severe 73 9: 10: lone-parents: 30); item [i] comments supra) (except IE MT) part dominate pursues WP RSWP (superscript population) write subgroups: countries/ regions): Thus regions) by: (1st 6); (2nd (3rd (4th 6) 11: RELATIVE SEVERITY diagonally striped contributing equation) Visual inspection reveals minor result; carries heavy weight Corluy (2013a indices’ inspired Gregg Scutella Wadsworth (2008 2010) ‘jobless household’ so-called (a jobless adult weeks survey) polarisation index actual hypothetical assuming distributed randomly about expected ‘individual intensities’ matched random formation Introducing random’ counterfactual sociological mathematically easier binary ‘jobless/non-jobless’ Yet employ Traditionally extended model negative lived structure) Polarisation became issue British agenda end nineties onwards declining 2000s Overall 1995 convergence EU: initially tended constitutes exception moving mid 2000’s (2013a) analyse stratification (2013b) dig long-term (space forbids pursuing here) enriched inquiry influencing decisions (14) 2008) 80; graphical 77 SOCIAL SPENDING cannot examining How characterise perspective? brings retrieved 2011:15 excluding (indicated ‘T-kids’); (P-kids); (T-all); (P-all) ‘Household income’ standardised usual equivalence scale take account; way (15) common practice administrative public published ESSPROS classification gauge importance well-known problem gross taxation regime benefits; tend overestimate real benefits like France (Adema Chart I 11) allow us assess demographic however thoroughly GDP 1) 78 12: Household All according OECD modified 79 allows spending: benefiting cash T-kids + P-kids ranked children’s biggest spender 16 pension P-all (P-all T-all) assessed solid dark grey Some Portugal ‘pension heavy’ Today Third orientation (both pensions) (T-kids P-kids) (T-all P-all) bar households) obvious variety respect: pension-heavy receive support Spending oriented’ (so conceived) Fourth (where amounts 50%) Slovakia 3%; supporting (16) spenders non-pension previous (17) de facto early retirement influence (18) illustrated report Employment Developments 2012 ‘poverty reduction pensions’ multigenerational seems (European Commission 222) underestimate emphasise hit spend unemployment become appear generous se doing underscored ante framework oriented tilted towards nonetheless cautious quality prima facie neither nor seem thing effectiveness ‘per euro’ ‘efficient’ fighting notoriously macro-level ‘efficiency’ strict Paretian word quasi-impossible efficiency spending’ Lefebvre Pestieau 16-18) conceptually ambitious ‘efficiency scoreboard’: Pareto-efficiency productivity conditional ‘inputs’ ‘pro-poorness’ conclusive still leaves substantial ‘unexplained disparity’ fabric architecture capita human capital readily ‘explained’ separately other) macro-comparative control scoreboard’ improves somewhat neutralised becomes Continental type scoreboard discounting bad ‘mediocre’ far excellent societal unambiguous practical certainly room improvement existence good Can improved? Micro-simulation helpful Maréchal Perelman Tarantchenko Van Camp (2010) allowances Up categories beneficiaries unemployed disabled pensioners) entitled supplements simulated reforms extend income-poor) tested qualify ‘modest’ attain amount people edge passing benefits): Simultaneously reach considerable baseline scenario (no reform) kind 70% reaches 97% Interestingly budgetary budget roughly 2% By 5th reform reduces extending 96 cost-effective operation ‘room improvement’ current extends incentive make transition inactivity (As supplement means-tested proposed create ‘income pay At sight worry central Admittedly studied couples single They (19) thank Guy giving scenarios 82 ‘work-rich’ miracles revisiting necessary broader re-examination dimension beyond domain standard earner increasingly determines decent dilemma whatever Next targeted creation activation) constellation replacement constrained disincentives (notably minimum wages ‘glass ceiling’ called Cantillon Mechelen (2014)) considerations cost competitiveness insufficient protect against wage parent 21 especially traps’) putting reconsider need schemes designed alleviate costs’ facing singles single-income earners child-rearing healthcare housing toolset greater assigned ‘cost-compensation’ intelligent nuanced principles selectivity today maximum billing care) development design services utmost (20) partially (2010): lone (21) Bogaerts Vandelannootte 9% paper: 100% 83 DISQUIETING DEVELOPMENTS: INTERTEMPORAL CHANGES adds evolution older surveys 2004 13: ELDERLY TOTAL POPULATION EUROSTAT hazardous instability SILC-based until Compared succeeding decreasing elderly choose fair trends consistent stable conceals intergenerational risks: rate; clear change); Over (22) Assessing starts from; den Bosch De Vil 2003-2011 optimistic 84 thresholds) policy-makers complacent agency Kind en Gezin collects fine grain [0-3] ‘kansarmoede’ encompassing stimulation register upward urban contexts findings warning [0-17] yet soon provided migration figures: mother non-Belgian origin ‘kansarm’ versus origin; growing markedly (Kind 2013) includes non-monetary non-material aspects exclusion partial data) integration migrant 14a cut-off evolutions broken down pre- post-crisis 14b 14c ‘Changes (keeping shares constant) Netherlands Czech Republic Flanders) contributed deteriorating 14A: BETWEEN 14B: 86 14C: ‘gain’ improving work-intensities Between dynamics 14c: boom Member States gain offset vulnerable followed interpreting forget boom) changes’ positive gives detail concerns group: increase23 Consequently half criterion 47 42 15: CHANGE 2006-2011 WITHIN (2006 evolved population; already born (23) formulate cautiously (24) recalculate entire 88 outside EU25 (that EU27 Romania) factor: (less ‘non-EU25’ everything Both non-EU25 100-102) participate generate native (SILC 2006-SILC coincided prosperity deliberate activating strategy vis-à-vis Should activation failure old distributive new Lisbon Strategy March 2000 deny tensions trade-offs soul-searching case: banner active welfare’ incentives incrementally improved lowering personal contributions bottom taxes earned close monitoring harsh sanctions foregoing summary possible prior challenges: traditional ‘financial incentive’ arsenal ill-equipped up Making bigger requires (to reduce unqualified leavers) possibilities existing stock low-skilled labour) calls targeting lone-parents) well-designed facilities affordable women earnings mothers) enhanced damaging introducing selective 89 CONCLUSION: POLICY CHALLENGES CONTEXT PERSISTENT DUAL POLARISATION Looked ‘haves’ ‘have nots’ consequences think region; complicated countries) Child youngest generation forecast trade-off doomed Activation traditionally organised recipe ‒ contrary ‘activation turn’ suggest went worsening redress enhance opportunities analysis; ‘work-rich standard’ characterises creates Possibilities 90 examined Together priority government 91 REFERENCES Adema W Fron Ladaique M Welfare State Really More Expensive? Migration Working Paper No 124 F (eds Poverty Reduction successful states? Oxford Press 2014 N Cracks paradigm security: International Review Sociology Vol 270-290 V Individual Methodologies Papers edition Union doi: 2785/41846 CSB-Working 2013b K Goedemé Vanhille J Union: 60-93 R workless Theory evidence Britain Australia Journal Population Economics 23 139-167 Two sides story: measuring Royal Statistical Society: Series (Statistics Society) 171 857-875 Het Vlaanderen L’Etat-Providence Performance Dumping Paris Editions Rue d’Ulm 92 C S E G Conference Presidency Council Marche-en-Famenne Sept Marx In-work 131-156 Active Revisited Bruges Die Keure Excessive Imbalances Euroforum KULeuven Mapping 1-59 evolutie van armoede bij ouderen nader bekeken Federaal Planbureau 6-13 Vandelannoote D Supplementary local agencies traps Flemosi April 93 APPENDICES APPENDIX DIFFERENCE RISK 5) (cutoffs: 20-45-55-85) categories) intensities; group; 94 A1: largest smallest consideration 95 TABLE (SUPPORTING Country AROP Explanatory Difference Sum Residual Share VLWI [0-0 2] [0 2-1] 97 A2 divide comment A2: 98 CONTENTS