The working poor, low wages and mobility out of poverty: A crosscountry

The working poor, low wages and mobility out of poverty: A crosscountry perspective Henning Lohmann University of Cologne LoWER Annual Conference European Low-wage Employment Research Network 15/16 April 2005, ZEW, Mannheim/Germany

1. Introduction: Work and poverty

1. Main topics of paper relationship between work and poverty: number of working poor employment entry of workless poor poverty exit via employment differentiating between low-wage, high -wage and self-employment 1

1. Working poor Few advanced countries, with the possible exception of the United States are confronted with rampant poverty in work, even countries where a fairly large fraction of the workforce is low-paid, like in Canada or the United Kingdom. (Marx/Verbist 1998: 80). Potential risk groups: single-earner households, low-wage workers, large households (see also O Connor/Smeeding 1995, Strengmann-Kuhn 2003) 2

1. Employment transitions and poverty exit Poverty dynamics: increase of earnings from employment most frequent route out of poverty (see e.g. Bane/ Ellwood 1986, Duncan et al. 1995, Layte/ Whelan 2003) Exit from welfare: exit from welfare via employment does not equal exit from poverty (see e.g. Buhr 2002, Cancian et al. 2002) Exit from low-wage employment: relevant share of low-wage workers stays in low-wage work over time (see e.g. Keese et al. 1998, Stewart/Swaffiled 1999, Cappelari 2002) 3

2. Data and indicators

2. Data and indicators User Data Base (UDB) of the European Community Household Panel (ECHP) observation period: 1994-2001 (most countries) 13 countries, EU before Eastern enlargement excluding Luxembourg and Sweden mainly regarding population in working age (17-64 years) cross-sectional analysis (2001), analysis of transitions (t t+1, pooled 1994-2001) 4

2. Poverty and low-wage employment poverty - relative income poverty - threshold: 60 percent of median monthly net equivalised household income (non-modified OECD scale) poverty exit - defined as being poor at t and being not poor at t+1 - income changes smaller than 5 percent of median income are not counted as poverty exit low-wage employment - threshold: 67 percent of median gross hourly wage (based on wages in current month, at least 15h) 5

2. Employment patterns time t t+1 workless workless low-wage low-wage high -wage high -wage self-employed self-employed 6

3. Working poor and workless poor

3. Working poor and workless poor Table 1: Poverty rates by employment status (working age population) total employment status low-wage 'high'-wage selfemployment not working employment employment Austria 7.5 11.8 10.2 3.1 15.3 Denmark 8.1 18.6 8.5 2.6 11.3 Germany 11.2 17.7 13.7 4.1 10.2 Belgium 11.6 17.9 11.1 5.4 6.3 UK 12.0 25.2 12.6 3.6 19.9 Finland 13.5 28.0 20.5 4.1 14.6 Portugal 13.6 22.2 14.8 6.4 16.5 Netherlands 14.0 18.9 14.8 8.7 17.6 Ireland 15.1 27.6 8.3 4.7 15.1 Greece 15.7 21.0 15.8 4.9 20.3 Spain 15.8 25.4 15.5 4.5 14.8 France 16.4 24.4 19.0 6.8 21.1 Italy 18.6 27.4 26.2 8.7 12.2 Ø 13.3 22.0 14.7 5.2 15.0 Source: ECHP 2001 (weighted), own calculations, Base: age 17-64 years, Ø: unweighted average 7

3. Working poor and workless poor Table 2: Distribution of workless and working poor (working age pop.) not low-wage 'high'-wage selfemployment working employment employment Source: ECHP 2001 (weighted), own calculations, Base: age 17-64 years, Ø: unweighted average total Austria 50.9 7.5 21.7 19.9 100 Denmark 66.3 6.0 20.2 7.5 100 Germany 66.3 11.3 16.5 5.9 100 Belgium 63.3 7.2 24.8 4.7 100 Finland 66.2 8.3 16.2 9.3 100 Netherlands 47.1 11.8 34.8 6.2 100 UK 61.1 11.9 14.5 12.5 100 Ireland 73.3 4.7 13.4 8.5 100 France 65.8 8.8 18.2 7.2 100 Spain 70.6 7.8 10.5 11.0 100 Portugal 51.5 6.0 23.1 19.4 100 Italy 68.5 5.3 17.0 9.2 100 Greece 56.9 6.1 9.4 27.6 100 Ø 62.1 7.9 18.5 11.5 100 8

3. Employment entry of workless poor Table 3: Employment entry of the workless poor (working age population) entry distribution of employment status if entry=yes low-wage 'high'-wage selfemployment no yes employment employment all entries Austria 81.8 18.2 24.7 56.5 18.8 100 Denmark 79.9 20.2 29.1 68.3 2.5 100 Germany 82.6 17.4 42.4 43.9 13.7 100 Belgium 90.7 9.3 28.6 63.9 7.6 100 UK 82.2 17.8 46.1 38.9 15.0 100 Finland 77.8 22.2 31.1 61.8 7.1 100 Portugal 79.6 20.4 27.4 47.1 25.5 100 Netherlands 83.5 16.5 54.3 36.6 9.1 100 Ireland 86.0 14.0 45.8 44.8 9.3 100 Greece 84.2 15.8 29.4 28.1 42.5 100 Spain 83.4 16.6 40.4 41.1 18.6 100 France 86.7 13.3 50.9 42.9 6.2 100 Italy 89.4 10.6 29.4 46.7 23.9 100 Ø 83.7 16.3 36.9 47.7 15.4 100 Source: ECHP 1994-2001 (weighted), own calculations, Base: age 17-64 years, Ø: unweighted average 9

4. Poverty exit

4. Poverty exit by employment pattern exit rate 0.2.4.6.8 1 PT IT FIN FR BE IRL GR UK GER ES NL AT DK country not working low-wage empl. 'high'-wage empl. low-wage entry 'high'-wage entry 10

4. Determinants of poverty exit discrete-time event-history model on probability of poverty exit risk set: workless poor (first observed poverty spell, no left-censoring) regarding the influence of employment patterns controlling for employment patterns of household members, demographic events, duration in poverty, sex, education, age (and country) 11

4. Determinants of poverty exit Table 4: Model on probability of exit from first observed pov. spell 1 empl. pattern t --> t+1 (ref.: not working --> not working): log odds s.e. low-wage empl. --> working 0.705 0.060 'high'-wage empl. --> working 1.017 0.045 self-empl. --> working 0.636 0.047 not working --> low-wage empl. 1.115 0.080 not working --> 'high'-wage empl. 1.696 0.079 not working --> self-empl. 0.639 0.110 working --> not working 0.146 0.069 no of observations 26820 no of events 12962 LR chi2 5628.8 df 38 Source: ECHP 1994-2001 (weighted), own calculations, Base: age 17-64 years 1) discrete-time event-history model, non-left-censored spells only, extract from larger model 12

4. Determinants of poverty exit by country 4 log_odds. 3 2 1 0-1 1 2 3 4 5 6 7 DK NL B FR IRL IT GR ES P A FIN D UK working entry exit Legend (employment patterns t t+1): 1: low-wage employment still working, 2: high -wage-employment still working, 3: self-employment still working, 4: not working low-wage employment, 5: not working high -wage employment, 6: not working self-employment, 7: working not working 13

4. Determinants of poverty exit summary of multivariate results: - the workless are least likely to exit from poverty - employment entrants are more likely to exit from poverty than those already working - likelihood of poverty exit by type of employment entry: high -wage > low-wage > self-employment - similar patterns for household members - weak and/or insignificant country differences - significant influence of demographic events (change of household size +, births -), age (U-shaped), education (+), duration in poverty (-) 14

5. Conclusion

5. Conclusion How large is the group of the working poor? - up to half of the poor are working - poverty risk of high -wage workers is rather low, poverty risk of low-wage workers is often above average 15

5. Conclusion Are the working poor more likely to exit from poverty than the workless poor? - probability of poverty is about 25 to 35 percentage points higher but it is low compared to employment entrants To what extent do transitions into employment lift the workless poor out of poverty? - the majority of high -wage entrants leaves poverty (on average ~75 percent), a bit less than half of the lowwage entrants stay poor 16

5. Conclusion Can we observe country differences? - weak differences in general patterns - differences in the probability of employment entry of workless poor: influences of economic and institutional framework (unemployment, labour market rigidities, welfare states)? - differences in the probability of poverty exit via employment: influences of economic and institutional framework (unemployment, bargaining system, welfare states)? 17