Do o older er f female e previou ous prison oner ers - PowerPoint PPT Presentation

Do o older er f female e previou ous prison oner ers participate in t the labour market? Joanna Felczak Ewa Gałecka -Burdziak Piotr Błędowski Marek Góra Warsaw School of Economics

Goals • Investigate whether employment experience protects from recidivism among older fe males who experienced a period of incarceration • Compare employment and unemployment spells • investigate if labour force participation differed in any manner before and after the imprisonment spell • verify if employment experience helped avoiding or postponing the subsequent incarceration spells

Sample (1) • administrative data from Poland on a group of older female workers (born between 1940 and 1965) who registered with any public employment office anytime between 1 June 2004 – 31 December 2017 and at t his time experienced one or more incarceration spells . • In this manner, we had a very specific, but precise sample of old-age first ever convicted . • The conducted research may trigger subsequent comparisons across birth cohorts, genders and/or countries

Sample (2) • Background : In Poland women make up less than 4% of all inmates. • They often break the law because of life challenges such as poverty, unemployment, and significant physical or behavioural health struggles, including those related to past histories of trauma, mental illness, or substance use. • Most of the incarcerated women had been previously, if ever, employed in low wage, entry-level positions ; many had received public assistance . Type s of performed occupation: • In 89.7% these were routine manual jobs, • in 6.7% routine cognitive, • 2.5% non - routine cognitive personal, • and in total 1% for non - routine cognitive analytical and manual physical.

Sample (3) • Entire sample 1,709 females Out of which: • 1,519 experienced exactly one incarceration spell, • subsequent 157 two such spells, • and the rest (33) at least three spells . They are disadvantaged fraction of the potential labour supply: older female workers, stigmatized and excluded due to incarceration - still young enough to be able to participate in the market and not old enough to be eligible for pension benefits . From individual and society perspectives it is expected and needed that these females successfully reintegrate with the labour market .

Table 1. Descriptive statistics of the individuals’ characteristics Mean Std. dev. Age at the beginning of the 1 st spell (in years) 46.0 4.4 Educational level (percentage distribution) primary 0.607 - vocational 0.195 - secondary vocational, general secondary, post-secondary and tertiary 0.199 - Total tenure (in years) 11.3 9.4 Unemployment benefit duration (in months) 6.0 3.7 Notes: no. of observations (individuals) 1,709, no. of observations for education 1,705, no. of observations for total tenure 1,320. Source: own elaboration.

Table 2. Descriptive statistics of the duration of particular s pells(in months) Mean Std. dev. employment 6.6 10.8 unemployment 9.2 11.5 non-participation 6.7 11.2 incarceration 7.4 5.8 Notes: no. of observations: employment 1,917; unemployment 7,802; non-participation 8,567; incarceration 1,941. Source: own calculation.

Methods (1) • L ongitudinal analysis (social sequence analysis, multi - state models) • T rajectories of the workers since the 1 June 2004 till the end of 2017. Social sequence: • we identified the status in the labour market of each individual in each day of the observation period. • We centralized the outcome around the beginning of each individual’s first incarceration spell. • In this manner we can compare the labour market status before and after imprisonment. • The results are presented in stacked tempogram in Figure 1 and 2

Figure 1. S Stacked d tempo pogram of t the he sha hares o of workers i in n a pa particular s state i in n the he l labo bour m market in n the he observatio ion pe perio iod d pr prior a and a nd after t the he be beginn nning o of the he first inc ncarceratio ion spe pell (uppe upper pa part) num number of of obs bservations ns (lowe wer pa part) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% -5.5 years 0 5.5 years employment unemployment incarceration non-participation

Number of observations 1800 1600 1400 1200 1000 800 600 400 200 0 -5.5 years 0 5.5 years no of observations

before Figure re 2. re 2nd inc. Spel 2. Te Tempogram for thos pell) 100% 10% 20% 30% 40% 50% 60% 70% 80% 90% 0% 100 120 140 160 180 20 40 60 80 0 -3128 -3128 -2992 -2985 -2856 -2842 -2720 -2699 ose, -2584 -2556 -2448 -2413 -2312 , who hav -2270 -2176 -2127 -2040 -1984 -1904 -1841 -1768 employment -1698 ave more than one -1632 -1555 -1496 -1412 -1360 -1269 -1224 -1126 -1088 -983 -952 -840 unemployment no. of observations -816 before 1st inc. spell -697 -680 -554 -544 -411 no. of observations -408 -268 -272 -125 -136 e incarc 18 0 incarceration 161 136 304 272 447 rcera 408 590 544 733 ration period ( 680 876 816 1019 non-participation 952 1162 1088 1305 1224 1448 1360 1591 1496 1734 1632 1877 d (befo 1768 2020 1904 2163 2040 2306 fore 1st inc. spel 2176 2449 2312 2592 2448 2735 2584 2878 2720 3021 2856 2992 ell and

100% 10% 20% 30% 40% 50% 60% 70% 80% 90% 0% -4503 100 120 140 160 180 20 40 60 80 -4313 0 -4503 -4123 -4285 -3933 -4067 -3743 -3849 -3553 -3631 -3363 -3413 -3173 -3195 -2983 -2977 -2793 -2759 -2603 -2541 employment -2413 -2323 -2223 -2105 -2033 -1887 -1843 -1669 -1653 -1451 -1463 unemployment -1233 -1273 no. of observations before 2nd inc. Spell -1015 -1083 -797 -893 no. of observations -579 -703 -361 -513 -143 -323 75 -133 incarceration 293 57 511 247 437 729 947 627 1165 817 1007 1383 non-participation 1197 1601 1819 1387 2037 1577 1767 2255 1957 2473 2691 2147 2337 2909 2527 3127 2717 3345 3563 2907 3097 3781 3287 3999 3477 4217 3667 3857 4047 4237

Results and discussion (1) • Incarceration preceded in 45% by non - participation and in 50% by unemployment spells. • Incarceration followed in 11% by unemployment, and in 88% by non - participation. These non participation spells were very short . O n averaged lasted 1.6 months, but in 50% of cases just 10 days, and in 97% of cases were followed by unemployment spells. • Non participation after incarceration should be as short as possible . Earlier employment/incentive is better for reintegration . • 57% of females did not work even once during the entire observation period, and we observed them on average for 7.7 years (the median equalled 8.2 years).

Results and discussion (2) • Standard example: unemployed – short incarceration – again unemployed and incarceration make her difficulties worse . We observe a long period of unemployment after the first sentence for those who had a subsequent sentence . Those who started work earlier were not convicted again. • L onger sentence – suprisingly - chance to improve qualifications and adapt to the requirements of the labour market . • Receive s ocial assistance allowance on the grounds of unemployment - no incentive to seek work. • Prisoners paroled under individualized post - release conditions are often obliged to seek employment with the help of a probation officer. • These coefficients should be interpreted that employment postpone recidivism and may reduce the risk of returning to prison. ( cause -effect relationship in further research)

Policy recomendation • The policy recommendation is that state actions should be aimed at increasing the labour market attachment of the previous and present prisoners. • Even if someone receive social assistance allowance on the grounds of unemployment should also participate in social work and be socially active to fulfill social roles .

Thank you for your attention! Joanna.Felczak@sgh.waw.pl