Social Inequalities and Education Part 3 Welfare regime

Social Inequalities and Education Part 3: Welfare regime stratification and social mobility Louis Chauvel Site : www. louischauvel. org/socineqedu Email : louis. chauvel@uni. lu 1

Welfare Regimes and mobility 1. 2. 3. 4. 5. Static and dynamic inequality – Sorokin Stratification, mobility and welfare regimes Class intergenerational mobility Income intergenerational mobility The role of education – Jerrim 2015 2

1. Static and dynamic inequality – Sorokin The difference between static inequalities (stratification) and dynamic inequalities (mobility) * Static and dynamic inequalities * The role of education * Ethno/urban inequalities and discrimination 3

Pitirim Sorokin, Social Mobility (1927) …and: Social and Cultural Dynamics Cincinnati: American Book Company, 1937 -41. 4 vol “Social stratification means the differentiation of a given population into hierarchically superposed classes. It is manifested in the existence of upper and lower social layers. Its basis and very essence consist in an unequal distribution of rights and privileges, duties and responsibilities, social values and privations, social power and influences among the members of a society. ” (chap 2) “Since vertical mobility actually functions to some degree in any society, there must be in the “membranes” between the strata “holes, ” “staircases, ” “elevators, ” or “channels” which permit individuals to move up and down, from stratum to stratum. The problem to be discussed now is : What are these channels of social circulation ? Various Social Institutions Perform This Function. —Among them there are few especially important from our standpoint. Of these few, which may be in different societies or in the same society, at different periods, one or two are particularly characteristic for a given type of society. The most important institutions of this kind have been: army, church, school, political, economic, and professional organizations”. (chap 8) Full text at archive. org https: //archive. org/details/in. ernet. dli. 2015. 275737 4

Pitirim Sorokin and Social Mobility 1. The principal forms of social mobility are horizontal and vertical. Vertical mobility exists in the form of ascending and descending currents. Both have two varieties: individual infiltration and collective ascent or descent of the whole group within the system of other groups. 2. According to the degree of the circulation, it is possible to discriminate between immobile and mobile types of society. 3. There scarcely has existed a society whose strata were absolutely closed. 4. There scarcely has existed a society where vertical mobility was absolutely free from obstacles. 5. The intensiveness and the generality of vertical mobility vary from group to group, from time to time (fluctuation in space and in time). In the history of a social body there is a rhythm of comparatively immobile and mobile periods. 6. In these fluctuations there seems to be no perpetual trend toward either an increase or decrease of vertical mobility. 7. Though the so-called democratic societies are often more mobile than autocratic ones, nevertheless, the rule is not general and has many exceptions. Full text at archive. org https: //archive. org/details/in. ernet. dli. 2015. 275737 5

2 - Stratification, mobility and welfare regimes Walter Korpi and Joakim Palme, 1998 http: //www. louischauvel. org/korpiclasswelfare 2657333. pdf 6

Another important dimension : (in)equality of opportunities at birth Hierarchy + US ? ? ? UK? ? ? Brazil Italy Fluidity – (closure) Fluidity + (openness) Finland Sweden France Germany Hierarchy – 7

Another important dimension : (in)equality of opportunities at birth Hierarchy + l de mo ral US ? ? ? ibe L Fluidity + (openness) UK? ? ? Soc Finland -de mo crat dre am Sweden Brazil Ext ineq reme uali ty Italy ive t rva x nse ado Co par France Fluidity – (closure) Germany Hierarchy – 8

3 - Class intergenerational mobility Goldthorpe class scheme Homonyms : Goldthorpe, Erikson–Goldthorpe, EGP (Erikson–Goldthorpe –Portocarero), and CASMIN (Comparative Study of Social Mobility in Industrial Nations) typology. 9

Class mobility Goldthorpe class scheme Is Social Mobility Really Declining? Intergenerational Class Mobility in Britain in the 1990 s and the 2000 by Yaojun Li and Fiona Devine Sociological Research Online, 16 (3) 4, 2011 10

Class mobility Is Social Mobility Really Declining? Intergenerational Class Mobility in Britain in the 1990 s and the 2000 by Yaojun Li and Fiona Devine Sociological Research Online, 16 (3) 4, 2011 11

Main references for the methodology www. louischauvel. org/yuxie 1992. pdf www. louischauvel. org/torcheasr 2005 chile. pdf www. louischauvel. org/vallet 2001 fourtyyears 3323052. pdf www. louischauvel. org/pisati 522871. pdf 12

Father or Mthr Portugal N=5536 Child Fluidity = log odds ratios (up+mid / lower class intergenerational mobility) = ln (658 x 3519/(235 x 1124)) 1 = 2, 17 (+/- 0, 17) CI [Ln(odds ratio)] = Ln(odds ratio) +/- 2 ni, j S 13

ISCO-88 (COM) International Standard Classification of Occupations recode PL 050 PM 070 (23 11 24 21 12 22 =1) (1 13 31 32 33 34 41 42=2) (else = 3) , gen(cat catp); Upper cl Middle cl Pop cl 14

Father or Mthr Portugal N=5536 Child Fluidity = log odds ratios (up+mid / lower class intergenerational mobility) = ln (658 x 3519/(235 x 1124)) = 2, 17 (+/- 0, 17) Father or Mthr Netherlands N=5041 Child Fluidity = 0, 86 (+/- 0, 13) 15

Father or Mthr Portugal N=5536 Child Fluidity = log odds ratios (upper / lower class intergenerational mobility) = ln (658 x 3519/(235 x 1124)) = 2, 17 (+/- 0, 17) Father or Mthr Netherlands N=5041 Child Fluidity = 0, 86 (+/- 0, 13) 16

Hierarchy + Fluidity + (d 9/d 1) Fluidity – Source : 2005 EU-SILC (Community Statistics on Income and Living Conditions) microdata N=201. 315 Hierarchy – Fluidity = log odds ratios (upper / lower class intergenerational mobility) 17

Unidiff=> 18

clear all set mem 400 m use "http: //www. louischauvel. org/mobsoc 2005 extr. dta" gen wgt=DB 090 gen rev=HX 090 tabstat rev [weight = wgt], statistics( mean p 10 median p 90 ) /// by(COUNTRY) columns(statistics) save table catp, by(COUNTRY) contract catp COUNTRY, zero freq(count) findit unidiff count, row(catp) col(cat) lay(COUNTRY) effect(add) pattern(fi) 19

Hierarchy + Fluidity + (d 9/d 1) Hierarchy – Fluidity = unidiff Kappa coefficient Fluidity – 20

4 - Income intergenerational mobility Static and dynamic inequalities = Discuss the accuracy of the model • • May be we can accept more static inequality if it goes with more intergenerational mobility (=meritocracy? ) Anyway, we have a difficulty: www. louischauvel. org/corak_2012. pdf • • On occupations, static and dynamic inequalities are independent On incomes (Corak 2012) the results are rather different • => the “Great Gatsby curve” 21

“Great Gatsby” Curve R-sq = 0. 75 !! Corak, Miles. 2013. “Inequality from Generation to Generation: The United States in Comparison. ” Chap. 6 in The Economics of Inequality, Poverty, and Discrimination in the 21 st Century, edited by Robert S. Rycroft. Santa Barbara, CA: ABC CLIO. (slope close to 1) 22 http: //milescorak. files. wordpress. com/2012/01/inequality-from-generation-to-generation-the-united-states-in-comparison-v 3. pdf

https: //www. diw. de/documents/publikationen/73/diw_01. c. 564651. de/diw_sp 0926. pdf More inequality, more viscosity? intergenerational Mobility in Europe Louis Chauvel Anne Hartung louis. chauvel@uni. lu anne. hartung@uni. lu IRSEI Institute for Research on Socio-Economic Inequality University of Luxembourg 23

xt Great Gatsby Curve: “More inequality, less social mobility”? R-sq = 0. 75 !! (slope close to 1) (Andrews and Leigh 2009; Blanden 2013, Corak 2013, Torche 2015) l. Relatively little is known about the link between income inequality and intergenerational mobility (Jäntti and Jenkins 2013) l. Is intergenerational mobility indeed greater in more equal countries? Corak, Miles. 2013. “Inequality from Generation to Generation: The United States in Comparison. ” Chap. 6 in The Economics of Inequality, Poverty, and Discrimination in the 21 st Century, edited by Robert S. Rycroft. Santa Barbara, CA: ABC - CLIO. 24

25 Intergenerational mobility (IM) in economics (income elasticity) and sociology (fluidity) The Great Gatsby curve (elasticities) Fluidity of societies (social class mobility tables, unidiff)

State of the arts & gaps in the literature Skip if no time … Relative level of intergenerational mobility differs depending on the measure used: income, education or social class (Blanden 2013, Torche 2014, Breen et al 2016) l Relatively little work on the mechanisms behind (Solon 2004, Breen and Jonsson 2005, Corak 2013, Jerrim and Macmillan 2015) Doubts on the Great Gatsby Curve (GGC) l. Corak’s version is a patchwork of surveys and methods / need direct comparisons l 2 S 2 SLS overestimates the slope and lacks robustness (Jerrim et al 2014, Lefranc) l. Parents’ incomes are often estimated on the base pf education and occ class l. The method is very sensitive to income distribution/inequality The role of education for intergenerational income mobility: A comparison of the United States, Great Britain, and Sweden / P Gregg, JO Jonsson, L Macmillan, C Mood 2017 Social Forces 96 (1), 121 -152 26

Is GGC trivial? higher elasticity whan Gini is higher So we have to work on ranks but we lose tails… ` Skip if no time … Issues with currency-elasticity II with the formula a ln(p/1 -p) = ln(medianised income) where a = Gini (Fisk-Champernowne-Dagum distributions), we simulate changes in the Ginis for fathers and sons based on the PSID 2 SLS estimation of currency elasticity after rescaling of fathers’ and sons’ Gini Þ Currency elasticity (in dollars) is Gini-change-dependent Þ When sons’ Gini > fathers’ Gini, elasticity increases 27

Research questions Can we confirm the link between income inequality (Gini index) and intergenerational mobility net of structural changes (Even with Gini independent measures of mobility)? What is the difference between “economic” approaches (income based) and “sociological” ones (occupational class)? What about countries versus welfare states? 28

Methods Logit rank controls inequality And keep the power tails Condense if no time … Our approach: Logit rank transformation Let us introduce logitrank based elasticity (RE) Coded on stata’s subroutine of our abg ssc install abg Society as a set of ranked positions p on the segment ]0, 1[ We just have to replace the yearly / country logged incomes by the logit of p “Logit rank” (O’Brien, 1978; Copas, 1999, Chauvel 2014) l. Similar to log(PSI) (Positional Status Index, Rotman Shavit, Shalev 2015) l. Strong correspondence with incomes Fisk-Champernowne approximation of medianized incomes m => ln(m) = α logit(p) with α = Gini (Champernowne, 1953; Fisk, 1960; Dagum 1975) => We compute elasticity not on log-incomes but on logitranks This rank based elasticity is a fathers’ and sons’ Gini-neutral measure of mobility net of structural changes in fathers/sons distributions 29

Same method in the context of social gradient of health: Louis Chauvel and Anja K. Leist 2015 “Socioeconomic hierarchy and health gradient in Europe: the role of income inequality and of social origins International Journal for Equity in Health 2015 14: 132 http: //equityhealthj. biomedcentral. com/articles/10. 1186/s 12939 -015 -0263 -y Methods Models Multilevel random intercepts random slopes model (Rabe-Hesketh & Skrondal 2012) l. Random country-specific intercept ζ 1 j l. Random country-specific slope for parental background β 2+ζ 2 j l. DV: yij “logit rank” in the equalized disposable income distribution (post-tax, posttransfer) l. EV: yij “logit rank” of the score of socio-economic origins (parents occupational class and education level) ζ 2 j measures the logitrank elasticity from parents to kids when ζ 2 j is high, socioeconomic reproduction is strong Loglinear association model l. Log-multiplicative layer model (Xie 1992, Pisati, xxxx) 30

Methods Data and variables Condense if no time … Data: EU-SILC 2005 and 2011 l. Modules “intergenerational transmission of poverty/disadvantages” l 26 countries: AT BE CY CZ DE DK EE ES FI FR GR HU IE IS IT LT LU LV NL NO PL PT SE SI SK UK l 52 subsamples l. PSID 2001 -2007 are the 53 rd sample => “US 2005” EV: max. available information on parents’ background l. MCA Multiple Correspondence Analysis (Burt method) of education and occupation of father/mother (6 classes EGP scheme) l. Predict factor coordinates: dim 1 explains 88. 6% inertia l. We mimic the same for US-PSID data (head of household’s father’s education and occupational group) 31

Results The origin gradient in different types of welfare states Nordic ζ 2 j = slope U. S. Source: EU-SILC 2005/2011, PSID 2001 -2007 32

The origin gradient across EU countries ζ 2 j = slope Source: EU-SILC 2005/2011, PSID 2005. Notes: Best linear unbiased predictor (BLUP) based on random coefficient null model. Labels: lower cases - 2005 (e. g. uk), upper cases - 2011 (e. g. UK). 33

Logit-rank based GGC Europe and the US R 2=. 343 ζ 2 j value Source: EU-SILC 2005/2011, PSID 2001 -2007 34

Results Logit-rank based GGC (a) Europe and the US R 2=. 343 ζ 2 j value (b) excluding post-socialist countries R 2=. 570 ζ 2 j value Source: EU-SILC 2005/2011, PSID 2001 -2007 35

Log-income based GGC This overestimates the R 2 due to the trivial fact income gradient is higher if gini is higher (a) Europe and the US R 2=. 661 ζ 2 j value (b) excluding post-socialist countries R 2=. 773 ζ 2 j value Source: EU-SILC 2005/2011, PSID 2001 -2007 36

Results Social-class based GGC (unidiff) (this captures nothing…) (a) Europe R 2=. 012 ζ 2 j value (b) excluding post-so R 2=. 079 ζ 2 j value Source: EU-SILC 2005/2011, PSID 2001 -2007 37

Explained country-level variance of the three if no time … Skip approaches Source: EU-SILC 2005/2011, PSID 2001 -2007 38

5 - intergenerational mobility & education John Jerrim, Lindsey Macmillan; Income Inequality, Intergenerational Mobility, and the Great Gatsby Curve: Is Education the Key? , Social Forces, Volume 94, Issue 2, 1 December 2015, Pages 505– 533 https: //johnjerrim. files. wordpress. com/2015/02/jerrim_macmillan_2015. pdf 39

https: //johnjerrim. files. wordpress. com/2015/02/jerrim_macmillan_2015. pdf 40

41

42

6 - Conclusion • • • Knowledge improved massively over the last decades A lot is still to be done in this respect Forecast: times of extreme inequalities are back Social rigidities – social immobility – should come back too ? Let’s be prepared … 43