Fertility Effects of Aggregate Unemployment Christian Schmitt SocioEconomic

  • Slides: 34
Download presentation
Fertility Effects of Aggregate Unemployment Christian Schmitt Socio-Economic Panel Study (SOEP) 2006 Conference of

Fertility Effects of Aggregate Unemployment Christian Schmitt Socio-Economic Panel Study (SOEP) 2006 Conference of the European Panel Users Network Universidad Pompeu Fabra de Barcelona, 8 -9 May 2006

Overview (1) Low Fertility in Europe - Brief Introduction (2) Research Topic and Previous

Overview (1) Low Fertility in Europe - Brief Introduction (2) Research Topic and Previous Findings (3) Theoretical Foundations (4) European Labour Markets during the 90 ies (5) Data and Methods (6) Findings and Conclusion EPUNet 2006

Background I – Below replacement levels of fertility among most western European countries Figure

Background I – Below replacement levels of fertility among most western European countries Figure 1: Total fertility rate (TRF) in EU countries 2001 Source: EPUNet 2006 European Communities 2003

Background II Potential causes: 1) Increasing female labour-force participation 2) Acquisition of economic resources

Background II Potential causes: 1) Increasing female labour-force participation 2) Acquisition of economic resources prior to family formation (also linked to labour-force participation) • Result: Strong integration into the labour force vs. time needed for family formation. • Time as a scarce resource Ø in a situational as well as in a Ø Life-course related sense EPUNet 2006

Background III Ø Difficulties in combining career and family roles, especially for women Conclusion:

Background III Ø Difficulties in combining career and family roles, especially for women Conclusion: High opportunity costs of parenthood EPUNet 2006

Research Topic • Second part of a research project dealing with fertility effects of

Research Topic • Second part of a research project dealing with fertility effects of unemployment • Part one: Effects of inidivdual unemployment • Findings in brief (lowered opportunity costs vs. reduced economic backing • Part two: Effects of aggregate unemployment EPUNet 2006

Research Topic and Previous Findings • Unemployment rate as economic predictor of future labour

Research Topic and Previous Findings • Unemployment rate as economic predictor of future labour market opportunities & prospects • Previous research based mainly on macro level models of economic indicators affecting fertilty Ø Unemployment Rate, Average earnings, GNP as predictor of TFR EPUNet 2006

Research Topic and Previous Findings Disagreement in previous reserach: Ø Pro-cyclical effects: • Wilkinson

Research Topic and Previous Findings Disagreement in previous reserach: Ø Pro-cyclical effects: • Wilkinson 1973, Macunovich&Easterlin (1988), Macunovich (1995), etc. Ø Vs. Countercyclical effects: • Butz & Ward (1979) (=> Relation to female earnings and reference to the New Home Economics) EPUNet 2006 • Major Problem: Derivations of aggregate data without understanding of the underlying effects and correlations on the micro level

Theoretical assumptions I Theory of Action • Maximisation of the utility under limited resources

Theoretical assumptions I Theory of Action • Maximisation of the utility under limited resources • Given constraints and preferences • Gender specific structuring of resources and constraints Ø New home Economics (esp. Becker 1981, 1993) • Gender specific division of labour between household and market work with the woman specialising in household work in most cases Brown Bag

Theoretical assumptions II Rational Choice and the Life-Course-Perspective • Preferences and expected utility are

Theoretical assumptions II Rational Choice and the Life-Course-Perspective • Preferences and expected utility are affected by life course transitions and trajectories Ø Social Change affecting the Life-Course: Educational Expansion and the role of human capital investments Ø Example: women may reject the role of the sole homemaker and work full time in order to avoid human capital depreciation Brown Bag

Theoretical assumptions III Institutions as Tie between Life-course an Rational Choice • Welfare-state institutions

Theoretical assumptions III Institutions as Tie between Life-course an Rational Choice • Welfare-state institutions structuring the life-course (Mayer 1990) • Institutions structuring (expected) utility • Imminent rationality vs. life-course dependent rationality Brown Bag Ø Rational-Choice core (example: discount assumption) vs. life course rationality Ø Family formation as rational choice, given a causal development of the life course (appearing as irrationality)

Theoretical assumptions IV High unemployment rate: Brown Bag Ø Increase in labour market related

Theoretical assumptions IV High unemployment rate: Brown Bag Ø Increase in labour market related uncertainties, decrease or staqnation in earnings potential Coping strategies for this situation? 1)Improvement of labour market position to prepare for expected risks Ø delay in family formation 2) Family formation as alternative pathway Ø temporary labour market exit as coping strategy Ø Gender specific differentiation (childbearing and childbirth affecting “female” time resources only) and Ø Cultural differentiation (different childcare cultures [family vs. public care] and different supply of childcare institutions; different family support of social policies)

Theoretical assumptions III Ø Individual fertility decision structured by (aggregate level) information, predicting economic

Theoretical assumptions III Ø Individual fertility decision structured by (aggregate level) information, predicting economic perspectives Three major questions: 1) Which information is used by the individual ? 2) How does the information (of high unemployment, e. g. ) affect individual decisions? Ø Decision making under uncertainty and Ø Effects of Risk Aversion 3) How does the individual interprete the information Ø Different threat potential of high unemployment rates in Brown Bag partial labour markets

Welfare state typology as frame of reference • Cross national comparison based on Esping

Welfare state typology as frame of reference • Cross national comparison based on Esping Andersen‘s typology of welfare regimes (1990) • Countries included in the analysis: • • EPUNet 2006 United Kingdom (anglo-american liberal) Germany (continental conservative) France (continental conservative) Finland (scandinavian social-democratic)

Evidence from the macro level Figure 2: Unemployment Rate in EU Countries Source: OECD

Evidence from the macro level Figure 2: Unemployment Rate in EU Countries Source: OECD (2004) Note: All values in percentage points EPUNet 2006

Evidence from the macro level Figure 3: Gender Unemployment Gap Source: OECD (2004) Note:

Evidence from the macro level Figure 3: Gender Unemployment Gap Source: OECD (2004) Note: All values in percentage points EPUNet 2006

Evidence from the macro level Figure 4: Total Fertility Rate Source: Eurostat (2006) Note:

Evidence from the macro level Figure 4: Total Fertility Rate Source: Eurostat (2006) Note: All values refer to TFR-Indicator EPUNet 2006

Social policy settings I – Family related benefits • France and Finland: Encouragement of

Social policy settings I – Family related benefits • France and Finland: Encouragement of the combination of family and occupational attainment (paternity leave, extensive childcare system) • Germany: Encouragement of women to retreat from the labour market (lasting childrearing leave, low coverage of childcare institutions) • UK: Dual pressure: Limited financial aid and high cost of widely privatised childcare system EPUNet 2006

Social policy settings II – Unemployment related benefits Unemployment insurance: • Finland: Tolerant rules

Social policy settings II – Unemployment related benefits Unemployment insurance: • Finland: Tolerant rules of entitlement, high payments, increased payment for parents, 23 months • Germany: High payments, increased payments for parents, 4 to 32 months • France: Payment 4 to 60 months, below 60% of last net income • UK: Low flat rate for 6 months Transfers are ceased in the UK after 6 months, in D, Fin and F subsequent unemployment assistance EPUNet 2006

Design of the multivariate analysis • Application of event history methods • Piecewise-constant exponential

Design of the multivariate analysis • Application of event history methods • Piecewise-constant exponential hazard model • Time variant (measured in months) and invariant covariates: Ø h(t) = exp( t) exp( x + zt) • Process time starts with 16 th year of life (month 192, population at risk: 16 -44 years of age) Brown Bag

Data and Methodology I • Consideration of Transitions to first birth is considered (family

Data and Methodology I • Consideration of Transitions to first birth is considered (family formation) • Month of fertilty decision as relevant event (to account for a causality) • Utilisation of ECHP-data from 1994 to 2001 • Separate estimates by country and gender EPUNet 2006

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet 2006 Source: ECHP 1994 - 2001, own calculations n = 5. 668

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet 2006 Source: ECHP 1994 - 2001, own calculations n = 7. 341

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet

Data and Methodology II Figure 5: Transition to first-parenthood – Kaplan-Meier survival estimates EPUNet 2006 Source: ECHP 1994 - 2001, own calculations n = 9. 865

Data and Methodology II Aggregate Information on: EPUNet 2006 • • • GNP, Annual

Data and Methodology II Aggregate Information on: EPUNet 2006 • • • GNP, Annual unemployment (NUTS 0) Regional annual unemployment rate (NUTS 1) Gender specific unemployment rate (NUTS 1) Age specific unemployment rate (NUTS 1) Individual activity status (incl. individual UE and duration) Information on previous long-term unemployment Individual and partners income and benefits and transfers Information on education, relationship, housing, origin, etc.

Findings I Aggregate Indicators (no Covariates) p < 0. 10 (*), p < 0.

Findings I Aggregate Indicators (no Covariates) p < 0. 10 (*), p < 0. 05 (**) and p < 0. 01 (***) EPUNet 2006 Source: ECHP 1994 -2001, own calculations

Findings I Aggregate Indicators (Full Model w. o. Partner) p < 0. 10 (*),

Findings I Aggregate Indicators (Full Model w. o. Partner) p < 0. 10 (*), p < 0. 05 (**) and p < 0. 01 (***) EPUNet 2006 Source: ECHP 1994 -2001, own calculations

Findings I Aggregate Indicators (Full Model with Partner) p < 0. 10 (*), p

Findings I Aggregate Indicators (Full Model with Partner) p < 0. 10 (*), p < 0. 05 (**) and p < 0. 01 (***) EPUNet 2006 Source: ECHP 1994 -2001, own calculations

Conclusion I • Effects for partial labour markets: Ø No significant effects for Persons

Conclusion I • Effects for partial labour markets: Ø No significant effects for Persons in labour market segments with higher and lower unemployment risks: Ø In detail: No significant effects for Ø managers and senior officials, professionals or for Ø service and sales-persons and machine operators EPUNet 2006

Conclusion II • Widely negative fertility effects of high unemployment rates • Special case

Conclusion II • Widely negative fertility effects of high unemployment rates • Special case for Finland • Gender specific uniform effect direction • No converse effect direction for GNP and Unemployment rates! • Especially clear effects for detailed indicators => Individuals use differentiated information for family formation decisions EPUNet 2006

Conclusion III France: • No or slightly negative effects (especially if controlling for youth

Conclusion III France: • No or slightly negative effects (especially if controlling for youth unemployment) Ø Possibly related to the cultural traditions of work-family combination and the extensive public support for families EPUNet 2006

Conclusion IV Finland: • Positive effects of high unemployment rates Ø Special case of

Conclusion IV Finland: • Positive effects of high unemployment rates Ø Special case of extensively high unemployment rates in Finland in the early 90 ies Ø Possible interpretation: Family formation as a focus on an alternative life goal (instead of labour market integration) Ø Backing by an extensive public support EPUNet 2006

Conclusion V Germany: • Diminished likelihood of family formation under high unemployment Ø Work-family

Conclusion V Germany: • Diminished likelihood of family formation under high unemployment Ø Work-family conflicts especially dominant in Germany (breadwinner-homemaker model) Ø Sequential model of family formation as common strategy (continued delay of family formation) Ø High level of social security shaping risk-averse behaviour EPUNet 2006

Conclusion VI United Kingdom: • No significant effects of aggregate economic indicators Ø Deregulated

Conclusion VI United Kingdom: • No significant effects of aggregate economic indicators Ø Deregulated labour market with a high level of entry and exit rates Ø Possibility: Labour market related risks are common and thus do not affect childbirth decisns. Ø Low level of social security, that diminishes subjective risk perception EPUNet 2006