WEBVTT

1
00:00:00.001 --> 00:00:08.000
Okay, would you like me to start? Okay. Well, good morning. The title of my paper today is the Intergenerational Reproduction of Self Direction revisiting class and conformity.

2
00:00:08.000 --> 00:00:15.000
You

3
00:00:15.000 --> 00:00:20.000
Yeah.

4
00:00:20.000 --> 00:00:21.000
Alright.

5
00:00:21.000 --> 00:00:24.000
I like to acknowledge my co-authors, Casper Burger. And Francisca Melee from the University of Zurich and Shaun and Han from the University of Minnesota.

6
00:00:24.000 --> 00:00:38.000
Now, generations of social scientists have studied the processes of social class reproduction. Parents influence their children's status attainment, including their income and occupational prestige.

7
00:00:38.000 --> 00:00:39.000
Okay.

8
00:00:39.000 --> 00:00:45.000
However, these attributes are not the only features of work that affect the quality of life.

9
00:00:45.000 --> 00:00:52.000
So.

10
00:00:52.000 --> 00:01:00.000
We can distinguish. Extrinsic. Let's see.

11
00:01:00.000 --> 00:01:10.000
Okay, so I'd like to just sandwich it. Extracts, extrinsic work rewards that status attainment researchers have mainly been concerned with.

12
00:01:10.000 --> 00:01:19.000
And these are rewards that you get for doing your work. They result from work. Income, occupational status and financial security.

13
00:01:19.000 --> 00:01:25.000
On the other hand, they're also intrinsic work rewards that are derived from work experience itself.

14
00:01:25.000 --> 00:01:26.000
Oh.

15
00:01:26.000 --> 00:01:34.000
Such as the opportunity to express one's interests, responsibility, and autonomy.

16
00:01:34.000 --> 00:01:40.000
And intrinsic work rewards have been given a lot of attention and studies of vocational development and career choice.

17
00:01:40.000 --> 00:01:50.000
The possibility that intrinsic rewards are transmitted intergenerationally has been neglected, however, in sociological theorizing and research.

18
00:01:50.000 --> 00:02:00.000
The transmission of intrinsic work rewards. Why be considered a potential mechanism through which any quality persists across generations.

19
00:02:00.000 --> 00:02:07.000
In this paper, we revisit Melvin Cone's foundational research on occupational self-direction.

20
00:02:07.000 --> 00:02:18.000
Cone hypothesized that self-directed work is transmitted from parents to children. This transmission process contributes to the reproduction of inequality.

21
00:02:18.000 --> 00:02:27.000
Cone posited 3 dimensions of self direction at work, closeness of supervision, the scrutinization of tests and the complexity of work.

22
00:02:27.000 --> 00:02:36.000
He reasoned that if your work is closely supervised and if it's very routine and repetitive, you can't be self-directed.

23
00:02:36.000 --> 00:02:44.000
But the more complexity there is in your work the more it will be required that you'd be self directed.

24
00:02:44.000 --> 00:02:53.000
And class in conformity, Cone hypothesize that intergenerational transmission through the central medium of values and orientations.

25
00:02:53.000 --> 00:03:05.000
Cone discovered links between social class position and parental child rearing values and practices. Middle class parents prioritize children's responsibility and initiative.

26
00:03:05.000 --> 00:03:17.000
Working class parents, on the other hand, emphasize behavioral conformity to external authority. Cone positive that these differences resulted from occupational conditions.

27
00:03:17.000 --> 00:03:24.000
That are either conducive to or restrictive of the exercise of self-direction and work.

28
00:03:24.000 --> 00:03:31.000
So here's a very simplified, diagram that, illustrates Cone's predictions.

29
00:03:31.000 --> 00:03:38.000
So, we have parental social status and parental self-direction at work. It's the exogenous variables.

30
00:03:38.000 --> 00:03:48.000
Values and behaviors of parents. Child varying practices of parents and then influencing children's values and behaviors.

31
00:03:48.000 --> 00:03:57.000
And this model was confirmed in many of Cone's studies as well as in the research of those who succeeded him.

32
00:03:57.000 --> 00:04:03.000
But however, Cone was not able to address the, final variables in this causal chain.

33
00:04:03.000 --> 00:04:11.000
The status attainment of the child is an adult and the children's self-direction.

34
00:04:11.000 --> 00:04:24.000
At work. Now in our research, we are studying the quality of the parent child relationship. And children's intrinsic work values and self-efficacy.

35
00:04:24.000 --> 00:04:37.000
These are not exactly the same variables that Cone had in his model, but Cone, reason that Cone had in his model, but Cone, reasoned that, middle class parents and those who were more self directed in their work.

36
00:04:37.000 --> 00:04:50.000
Would, have higher quality relationships with their children, closer and more attentive because they would be more concerned with the children's internal psychological development.

37
00:04:50.000 --> 00:05:07.000
We're studying children's intrinsic work values and self-efficacy and these values are are these variables are very much implicated in self-direction and then we can study also the status attainment.

38
00:05:07.000 --> 00:05:20.000
Of the child and the children self-direction at work. So our key research questions are is there intergenerational transmission of occupational self-direction as well as socioeconomic status.

39
00:05:20.000 --> 00:05:31.000
And is the intergenerational transmission process mediated by the quality of parent child relationships and adult children's intrinsic work values and self-efficacy.

40
00:05:31.000 --> 00:05:44.000
We draw in data from the youth development study that has been going on now for more than 50 years. Starting off with a, group of more than a thousand ninth graders age 14 and 15 in St.

41
00:05:44.000 --> 00:05:58.000
Paul, Minnesota in the United States and they've been followed through midlife. We have data on this cohort as well as their parents and their adolescent children.

42
00:05:58.000 --> 00:06:09.000
However, for this particular paper, our analytics sample consists of just the G one mothers and fathers studied during the G 2 children's adolescence.

43
00:06:09.000 --> 00:06:20.000
Followed, through. 2011 when they were 37 and 38 and second generation.

44
00:06:20.000 --> 00:06:35.000
So this is a figure that, describes the, parables and they relationships that we're most interested in with self direction and a one So socioeconomic status.

45
00:06:35.000 --> 00:06:50.000
Hypothesized to influence the same variables in G 2 mediated by the quality of the parent-child relationship, intrinsic work values and the child's self-efficacy as an adult.

46
00:06:50.000 --> 00:07:02.000
As for our measures, we measure, self direction by asking questions about do you have to think of new ways of doing things or solving problems on your job?

47
00:07:02.000 --> 00:07:12.000
How much control do you have over the way you spend your time at work and overall how much freedom do you have to make important decisions about what you do at work and how you do it?

48
00:07:12.000 --> 00:07:20.000
So these questions were asked of both the G one parents and the G 2 children.

49
00:07:20.000 --> 00:07:29.000
The socioeconomic status of both generations is indicated by educational attainment and their occupational status.

50
00:07:29.000 --> 00:07:49.000
We measured the quality of the parent child relationship by asking the atolescent children about whether they shared enjoyable activities with their parents discussed important concerns and issues with them and how close they felt to their parents.

51
00:07:49.000 --> 00:07:55.000
Intrinsic work values are indicated by a chance to make my own decisions, a job where I have a lot of responsibility.

52
00:07:55.000 --> 00:08:05.000
And a chance to learn a lot of new things. Those who value those highly would be more self-directed.

53
00:08:05.000 --> 00:08:21.000
And then self-efficacy is indicated by items from the Pearl and Mastery scale indicating how much the adults felt that they could deal with the problems of their.

54
00:08:21.000 --> 00:08:25.000
No.

55
00:08:25.000 --> 00:08:31.000
Let's see.

56
00:08:31.000 --> 00:08:41.000
And we also, control race. Gender and family structure.

57
00:08:41.000 --> 00:09:02.000
Now these are our estimates and you can see here in that. As hypothesize self-direction and socioeconomic status of G one predicts the same variables in G 2 and there are also some interesting mediating pathways.

58
00:09:02.000 --> 00:09:10.000
I'd like to, you know, highlight some of these. Pathways. In this, presentation.

59
00:09:10.000 --> 00:09:36.000
So first of all, the for the direct pathways. We see that. One socioeconomic status affects G 2 socioeconomic status and the magnitude of the relationship is much stronger than for self direction but the path coefficient for self direction is still statistically significant providing some support for Cone's

60
00:09:36.000 --> 00:09:49.000
theory. We find that. The intervening variables in our model don't help us to explain the continuity of self-direction across generations.

61
00:09:49.000 --> 00:10:00.000
But we do find some interesting mediating pathways implicating the relationship between G one socioeconomic status and G 2 self-direction.

62
00:10:00.000 --> 00:10:14.000
So you see here that G one, SES, influences the quality of the parent child relationship. The adult child's intrinsic values and self-efficacy.

63
00:10:14.000 --> 00:10:27.000
And then from there the self direction of the G 2 adult. And then not counting the parent child relationship, the G one SCS also.

64
00:10:27.000 --> 00:10:36.000
That's directly, intrinsic work values and self advocacy which in turn influence self-direction of J 2.

65
00:10:36.000 --> 00:11:00.000
Now turning to the transmission of socioeconomic status, we find again that G one socioeconomic status influences the quality of the parent child relationship, which in turn affects intrinsic values and G 2 socioeconomic status as well as more sort of direct relationship that doesn't involve the

66
00:11:00.000 --> 00:11:10.000
parent-child relationship quality. So we find here that there's evidence of transmission of self-direction at work from parents to children.

67
00:11:10.000 --> 00:11:23.000
But the intervening variables did not significantly contribute to this transmission process. Parental socioeconomic status indirectly influence children's self-direction at work via pair of child relationships.

68
00:11:23.000 --> 00:11:37.000
Intrinsic work values and self-efficacy. And, we find a sizable direct link between socioeconomic status across generations and indirect links by a family relationships and intrinsic.

69
00:11:37.000 --> 00:11:51.000
Work values. Now there are some limitations to our study. The first concerns generalizability. The data and findings mirrored the composition of the school population at the time the study was.

70
00:11:51.000 --> 00:12:02.000
Initiated so we can general relies to more recent cohorts or, the US at large or any other country of course.

71
00:12:02.000 --> 00:12:18.000
And we could not really address the entire occupational language hypothesis. The whole causal chain as cone described it Since the White House collected no G one measures of G one self directed orientations.

72
00:12:18.000 --> 00:12:30.000
Or G one child rearing values and practices. Still our findings extend the existing literature by documenting continuity of self-directed work across generations.

73
00:12:30.000 --> 00:12:46.000
Indirect links between parental socio economic status and adult children self-direction. Via family relationships and especially adult children's motivational resources, their intrinsic values and their self-efficacy.

74
00:12:46.000 --> 00:12:56.000
So in line with Cone's predictions, we find evidence suggesting work related socialization that extends beyond social class indicators.

75
00:12:56.000 --> 00:13:14.000
Thank you for your attention and I just like to point out that the youth development study data used in the present study as well as our other data are publicly available through the University of Michigan's Inter-university Consortium for Political and Social Research.

76
00:13:14.000 --> 00:13:16.000
Thank you.

77
00:13:16.000 --> 00:13:20.000
Thank you.

78
00:13:20.000 --> 00:13:26.000
So, let's see.

79
00:13:26.000 --> 00:13:41.000
Is no online question so we would So she needs to stop here. Oh, okay. So first we would need to stop sharing your slides here.

80
00:13:41.000 --> 00:13:52.000
And then, let's see if we have any questions in person here. In our room.

81
00:13:52.000 --> 00:14:07.000
And please use the microphone right in front so that everyone can hear you.

82
00:14:07.000 --> 00:14:16.000
I think based on the schedule in the program so we will save all the questions to the end after all the presentations.

83
00:14:16.000 --> 00:14:30.000
Sure. Since we started 5 min late So we are about to have the second presentation. Which is you.

84
00:14:30.000 --> 00:14:35.000
Yes, so my co-author Hunter York, recorded a video. Do you see that?

85
00:14:35.000 --> 00:14:38.000
Can you play it?

86
00:14:38.000 --> 00:14:49.000
We have, this shows that it is an online presenter rather than pre-recorded one.

87
00:14:49.000 --> 00:14:55.000
I think you should be able to see the video. So we can place a video. Cause he cannot be here.

88
00:14:55.000 --> 00:15:00.000
He's currently in France, so it's like midnight.

89
00:15:00.000 --> 00:15:08.000
Wait, that's not, I don't think that's a video or.

90
00:15:08.000 --> 00:15:09.000
Oh.

91
00:15:09.000 --> 00:15:14.000
Hi everyone, I'm Punjab Berk. I'm a PhD student at Princeton University.

92
00:15:14.000 --> 00:15:23.000
My presenting work co-authored with Yusha Princeton University of She's and she song, and I worked in tile and gradationalism revisited.

93
00:15:23.000 --> 00:15:30.000
It's on intergenerational intergenerational occupational mobility along axes of occupational characteristics.

94
00:15:30.000 --> 00:15:37.000
So in the literature there's generally 2 ways of organizing occupations. We all know of class models which are categorical.

95
00:15:37.000 --> 00:15:51.000
They use groupings to delineate occupations. But share similarities and so some examples are Marxian power relations or more with varying worker characteristics or alternatively just simply mobility chances.

96
00:15:51.000 --> 00:15:57.000
Here's a stylized representation of a class, categorical scheme. We have too late maxes.

97
00:15:57.000 --> 00:16:04.000
And then you have clearly, to donated groups, occupations. That exist in the space of these axes.

98
00:16:04.000 --> 00:16:10.000
Alternatively, you have gradational models which make no a priori presumptions about occupational groupings.

99
00:16:10.000 --> 00:16:19.000
And they assume occupations lie on one or more axes of subcontinuous indicator. And so here's what that would look like in this same fictional space.

100
00:16:19.000 --> 00:16:32.000
Where there's not clear breaks between the occupations that form classes. As was noted in the previous slide, classes sometimes overlay continuous ideas even when they're not mentioned as such.

101
00:16:32.000 --> 00:16:35.000
So we all recognize this for do, graph. Where he put like, occupations themselves and also, you know, lifestyle characteristics.

102
00:16:35.000 --> 00:16:58.000
In this latent space described by 2 axes. And so even though he talked about, you know, class fractions, you know underlying his work was this idea that there was a continuous space on which the social, space existed.

103
00:16:58.000 --> 00:17:18.000
Also from, arts classes we see that even though he was you know a Marxian and a categorical list underlying his scheme of classes there were these these continuous concepts of skill and credentials and organizational assets.

104
00:17:18.000 --> 00:17:27.000
Not to mention, ownership of the means of production. So all that to say, categoricalism has been a success for me.

105
00:17:27.000 --> 00:17:40.000
We all think of classes often. We use them to describe the real world. And this is because of gradualism often was just viewed as being too simplistic and so it fell out of favor in sociology.

106
00:17:40.000 --> 00:17:51.000
As a profession we love to make groups and often because of methodological limitations we're forced to work with groups even when we think things might be continuous.

107
00:17:51.000 --> 00:17:58.000
And so this is ultimately led to micro class theory as just like the pinnacle of categoricalism.

108
00:17:58.000 --> 00:18:10.000
It was born of these 2 earlier bullet points and more. Which was that to locate crisis at the site production themselves, air grow occupations.

109
00:18:10.000 --> 00:18:21.000
And underlying this theory is the idea that we are socialized both into and by occupations. So as a worker in a given occupation, my my fellow workers shaped me in some way.

110
00:18:21.000 --> 00:18:36.000
That results in some sort of social closure and some sort of like shared social conditioning. And this macro to my growth shifts that the micro class theory is part of is part of a larger end that privileges the role of smaller level of analyses.

111
00:18:36.000 --> 00:18:44.000
And chiefly the individual. And so this is related to the causal revolution and all these other things.

112
00:18:44.000 --> 00:18:55.000
And so where we're situating our work. Can be depicted by this four-dimensional sorry, 2 dimensional like 2, 2 by 2 cross tabulation.

113
00:18:55.000 --> 00:18:58.000
Of categoricalist and gradational. Perspective, the unit in few dimensional and multi-dimensional perspectives.

114
00:18:58.000 --> 00:19:10.000
So in the top left corner you have just simple three-class models. Egp, Mark, again, etc.

115
00:19:10.000 --> 00:19:19.000
You have their gradational equivalence, which are just SEI indices, status prestige, house model of SAT.

116
00:19:19.000 --> 00:19:23.000
And more in the multi-dimensional categorical perspective. You have microclasses and a few other.

117
00:19:23.000 --> 00:19:34.000
One off studies. And then in the gradational multi dimensional perspective there are very few studies.

118
00:19:34.000 --> 00:19:44.000
We found a few that kind of fit this perspective, Lebanon and Gresky. but none of them are exhaustive and none of them seek to describe occupations.

119
00:19:44.000 --> 00:19:51.000
A priori using these. They've used them mainly as just a byproducts of their analyses and they're not the main point of their study.

120
00:19:51.000 --> 00:20:00.000
So what we're proposing is a multi-dimensional gradational model. In the bottom right corner and I will sc that more now.

121
00:20:00.000 --> 00:20:09.000
So. The key innovations of our model. Is that off diagonal movement is not described by class theory.

122
00:20:09.000 --> 00:20:19.000
Least of all by micro class theories. So this would be upper class often, you know, could be it's upper class, but also in the Micro Cross example, doctors have doctors as sons.

123
00:20:19.000 --> 00:20:28.000
Etc. Off diagonal movement would be like nurses having doctors as children or farmers having agricultural scientists as children.

124
00:20:28.000 --> 00:20:39.000
I keep strength of, you know, being able to characterize off to act no movement with a gradual model is that it is generalizable across 10 years without explicit accommodations.

125
00:20:39.000 --> 00:20:45.000
So in the above examples Here's this beginning doctors, you know, these are gender type occupations and while they're not describeable with.

126
00:20:45.000 --> 00:21:04.000
Microsoft model they might be describeable with a more, gradational platform. And then on top of this, as I said earlier, This is not an attack on class thinking at all, especially because it allows categoricalism itself to be tested.

127
00:21:04.000 --> 00:21:13.000
And so if there's structured remaining after we account for gradational characteristics, like there is some, meet to class 3 and it's worth exploring more.

128
00:21:13.000 --> 00:21:21.000
Which we'll talk about more later. And so how are we doing this? We have a few sources of data.

129
00:21:21.000 --> 00:21:31.000
So our first set of data is we have, on the left these own at indicators, 244 of them basically all continuous indicators.

130
00:21:31.000 --> 00:21:52.000
From 8 domains in the own at database And so for every, SOC occupation, you have, you know, a measure from on a continuous scale of things like armhand steadiness and abilities you also have things like, work activities, context credentials.

131
00:21:52.000 --> 00:21:58.000
We have skills which people use most often, but there are also things like values which people have used less often.

132
00:21:58.000 --> 00:22:08.000
Things like achievement, independence and recognition. So what we use is an exploratory factor analysis to reduce 244.

133
00:22:08.000 --> 00:22:17.000
Columns into just 10 composite factors. Of which fibers shown. These have been given names.

134
00:22:17.000 --> 00:22:20.000
Manually But you can see that they match their top 5 components at their maximum and minimum extreme occupations quite a bit.

135
00:22:20.000 --> 00:22:27.000
I'll see that they match their top 5 components at their maximum and minimum extreme occupations quite a bit.

136
00:22:27.000 --> 00:22:34.000
So I want to draw your attention to our third most salient characteristic which is concern caringness and self-control.

137
00:22:34.000 --> 00:22:41.000
The top 5 components of this worth things like work styles of self-control and concern for others and social orientation.

138
00:22:41.000 --> 00:22:48.000
Work activities such as assisting and caring for others and values such as relationships. The extreme for this is nurses.

139
00:22:48.000 --> 00:23:02.000
Which are student professional nurses and at the minimum extreme is and so the reason why this is important is because you know often times We talk about skills in describing occupations.

140
00:23:02.000 --> 00:23:17.000
And we use Onet to measure them, but, by including work styles and values and work activities, we see that there there are other salient features of occupations that describe them.

141
00:23:17.000 --> 00:23:25.000
So you might call these soft skills in some domains, but you know, Onet doesn't even call them skills at all.

142
00:23:25.000 --> 00:23:35.000
Anyway, we have 10 of these, latent characteristics and they describe, you know, over 73% of variation in the data.

143
00:23:35.000 --> 00:23:43.000
We combine them with inter-generational occupational mobility data from 11 sources. These are basically exhaustive surveys.

144
00:23:43.000 --> 00:24:00.000
That include occupations of parents and children. We include all parent child gender pairs. We pull these data and they represent birth cohorts ranging from 1898 to 1993.

145
00:24:00.000 --> 00:24:11.000
Over 239 occupations. And over 15150,000 parent child pairs.

146
00:24:11.000 --> 00:24:20.000
To model our data, we use log linear models. These are important because they allow us to isolate structural and exchange mobility.

147
00:24:20.000 --> 00:24:27.000
That is to say we account for the baseline occupational dirt distributions of occupations in parents.

148
00:24:27.000 --> 00:24:33.000
And their children and then we see what movement happens in excess of that. We use nested models.

149
00:24:33.000 --> 00:24:41.000
There go, we compare the addition of class inability effects at varying levels with the addition of gradational mobility and inability effects.

150
00:24:41.000 --> 00:24:47.000
These models are direct antecedents of how it's status autonomy training model and Gresky and colleagues models.

151
00:24:47.000 --> 00:24:53.000
And oftentimes they're a combination of the 2.

152
00:24:53.000 --> 00:24:59.000
So what happens when we do this? I'm going to show you a huge table, which is very hard to explain.

153
00:24:59.000 --> 00:25:03.000
Gonna do my best. And so I'm just gonna draw your attention to AIC for now.

154
00:25:03.000 --> 00:25:11.000
And so this is a Goodness of fit, a measure of model goodness of fit. Lower numbers are better.

155
00:25:11.000 --> 00:25:18.000
We have for father's sons, father's daughters, mother's sons and mother's daughters.

156
00:25:18.000 --> 00:25:24.000
And the first thing I want to point out is that comparing 3 nested models each nested in each other.

157
00:25:24.000 --> 00:25:39.000
We add, the baseline model which has R calling marginal effects and SCI. We see that we can replicate Gresky's findings, which says that adding micro miso and micropass immability effects improves model goodness of fit.

158
00:25:39.000 --> 00:25:45.000
However, we also see that going beyond this including gradational effects. Further improves the fit.

159
00:25:45.000 --> 00:25:56.000
So that is 1 one argument for including gradual effects. If we just compare a purely gradational model and a purely categorical model head to head.

160
00:25:56.000 --> 00:26:01.000
We see that the gradational model out performs the categorical model. And so models 5 and model 7.

161
00:26:01.000 --> 00:26:18.000
The first one includes just ruin call marginal effects and just 11 gradational effects. Model 7 includes rowing card, column marginal effects and then over 80 categorical effects capturing micro miso and macro class immobility.

162
00:26:18.000 --> 00:26:35.000
We see that AIC, which is not adjusted for model complexity, already outperforms. They're in the gradationalist example When we use BISC, which accounts for model complexity, we see a further improvement in fit.

163
00:26:35.000 --> 00:26:42.000
Now I'm going to point out that the best model differs for each parent child gender pairing.

164
00:26:42.000 --> 00:26:52.000
But there's a general trend which is that, gradual effects matter. So in, father's daughters, mother's sons and mother's daughters, the best model.

165
00:26:52.000 --> 00:26:59.000
Was the model with purely gradational effects, so 11 gradational effects. And SEI, which is also gradational.

166
00:26:59.000 --> 00:27:06.000
So this would be 12 gradational effects. For further than sons, we actually do see that the Microsoft model.

167
00:27:06.000 --> 00:27:20.000
Fits quite well, but that the best performing model is the Microsoft model and the gradational effects model. And so this shows that there is occupational inheritance and then beyond direct occupational inheritance when people do not exist on the diagonal.

168
00:27:20.000 --> 00:27:32.000
It helps to explain their movement with gradational characteristics. So this is the full table that you take more time to look at and it just includes, like log likelihood ratio tests.

169
00:27:32.000 --> 00:27:39.000
And to show how, different combinations of models compare against one another.

170
00:27:39.000 --> 00:27:43.000
So now I wanna draw your attention directly to the microclass effects themselves. So this is that baseline.

171
00:27:43.000 --> 00:27:54.000
Gresky replication model from the previous slide and we see that many, many, many micropass effects are significant in a model.

172
00:27:54.000 --> 00:28:09.000
This means that, sons of ship officers will themselves often become ship officers. And so this is what Gresky, thought was, evidence in support of his model and indeed it is.

173
00:28:09.000 --> 00:28:34.000
However, when we add scaling effects we see that many many many of these microquest effects fall out and so after adding just 11 scaling effects we see that the bulk of the significant micro class inheritance effects become and and and this also happens in father's daughters and other sons and mother's daughters even though they were

174
00:28:34.000 --> 00:28:41.000
less. And this also happens in father's daughters and mother's daughters, even though there were less, that were significant in the first place And here they are compared side by side.

175
00:28:41.000 --> 00:28:56.000
We can also look at the scaling effects themselves. So these are the gradational measures. These were the, 10 or 11 gradational effects I was talking about earlier.

176
00:28:56.000 --> 00:29:09.000
And so the important thing here is that for most of them, they are both positive and significant, which means that there is a inheritance of gradational effects beyond just simply inheriting.

177
00:29:09.000 --> 00:29:31.000
Once occupation so these are specifically gradational effects which describe off to act on movement some of them are negative and significant, which, we account for by, either age biases in our data or Just changes in the occupational structure over time that can't be captured by a lot of linear model.

178
00:29:31.000 --> 00:29:48.000
And so like for example, for the finance and management. Case the way a lot of these surveys work is they simply ask an adult what was their father what was his occupation and your you know when you're asked that you're maybe 30 years old but you're likely to remember your dad's occupation when he was

179
00:29:48.000 --> 00:29:52.000
older when you were a child's occupation when he was older when you were a child or even your dad's occupation when he was older when you were a child or even now if he's still working.

180
00:29:52.000 --> 00:30:03.000
And so that biases it towards management occupations simply because the older you are, the more you move up, management axis

181
00:30:03.000 --> 00:30:07.000
Inability effects are more of a mixed bag. For fathers and sons, most of them are positive and significant.

182
00:30:07.000 --> 00:30:20.000
There were there are some negative and significant ones. This actually replicates my Callous SAT model where he found that the S the status and really the effect was negative.

183
00:30:20.000 --> 00:30:31.000
He comes, he and we, agree that this might be explained by, Again, shifting occupational characteristics.

184
00:30:31.000 --> 00:30:38.000
Or just distributions where for instance like there are large outflows of certain occupations into other occupations.

185
00:30:38.000 --> 00:30:45.000
Like farmers becoming something else. That accounts for basically these, effects.

186
00:30:45.000 --> 00:30:53.000
And so that sums up our results. The implications of this model is that there are in just general sociology.

187
00:30:53.000 --> 00:31:00.000
It allows for more robust and interpretable occupational class controls without using microbiases. And so you can imagine.

188
00:31:00.000 --> 00:31:13.000
If you want to add fixed effects, to your model. But don't have that many degrees of freedom to work with you could use these to kind of proxy occupational classes or occupations themselves.

189
00:31:13.000 --> 00:31:22.000
Theoretically, this begs the question of if class like structure is not happening at the level of occupations, where is it happening if it is happening and how is it happening?

190
00:31:22.000 --> 00:31:36.000
And so those would be rigidities that are left in excess of our model. And then for the intergenerational mobility literature, this recognizes, you know, the potential trade-offs individuals make between vertical and horizontal mobility.

191
00:31:36.000 --> 00:31:49.000
So, you know, Wilmer says written about the pro-sociality penalty. Which is to say that you know, even though we're describing what you might describe as horizontal, occupational stratification.

192
00:31:49.000 --> 00:32:00.000
If people are more inclined to, inherit the horizontal characteristics of their parents, it might come with vertical penalties, which are worth setting.

193
00:32:00.000 --> 00:32:11.000
And the takeaways. Of our study or that in our data the occupational structure exists in a space that can be sufficiently described in just a few dimensions.

194
00:32:11.000 --> 00:32:24.000
These dimensions are composites of activity, skills, knowledge, lifestyles and values. And they explain an amount of variation on par which is what explained is explained by class reproduction and men and even more so in women.

195
00:32:24.000 --> 00:32:34.000
This happens because they can explain movement within causes and movement between classes and occupations for the off diagonal event.

196
00:32:34.000 --> 00:32:41.000
And we do not say that class is dead, but we should really, really reconsider where we're looking for it.

197
00:32:41.000 --> 00:32:45.000
And so if it's not happening at the site of occupations for happening. Perhaps it's happening.

198
00:32:45.000 --> 00:33:04.000
In aiances that form kind of between occupations. Kind of at the miso level but like not in the miso paradigms that have been proper before so it could be worth while exploring you know a downwards up kind of perspective and making a new class model, exploring, you know, a downwards up kind of perspective in making a new class model if that's what

199
00:33:04.000 --> 00:33:10.000
we're So thank you. I would love to hear your comments, so please email me at H York at Princeton.

200
00:33:10.000 --> 00:33:31.000
EDU or go to my website and I'd like to also thank my co-authors and thank everyone who organized this conference.

201
00:33:31.000 --> 00:33:49.000
Okay, thanks. And then we have the third presentation in person.

202
00:33:49.000 --> 00:34:00.000
Thank you all for coming to this morning session. Before you leave for the conference. You still come here.

203
00:34:00.000 --> 00:34:16.000
Let me say a few words about the papers background. This paper is actually 2 papers in one. Originally designed to be 2 papers, I folded into one to make it the lead paper of a set of about 10 paper projects at different stages.

204
00:34:16.000 --> 00:34:28.000
So a lot of stuff in it. I don't want to overload everyone. Is what I would just focus on the substantive story.

205
00:34:28.000 --> 00:34:51.000
But let me begin by saying a few things about the hidden methodological paper behind the hope. Presentation. The methodological paper You actually the most sampling contribution of my contribution to my beloved RC, 28.

206
00:34:51.000 --> 00:35:02.000
I've spent 3 decades at OCI. 28 learn a lot and love this committee a lot and of all the presentations I've made this one is the most senator.

207
00:35:02.000 --> 00:35:13.000
And I dedicate the paper, the methodological idea to the memory of Leo Dudman. Oh, was my teacher when I was still a pixie student at Chicago?

208
00:35:13.000 --> 00:35:27.000
He left. Soon after I joined Chicago then. He introduced me to association modeling. That was like, 40 years ago.

209
00:35:27.000 --> 00:35:39.000
And I rarely use it except in one. Paper when I was a graduate student I never returned to using it because of an unuse of the whole approach.

210
00:35:39.000 --> 00:35:46.000
This is the hidden elephant in the room. We didn't talk about it at C. March.

211
00:35:46.000 --> 00:35:54.000
Except for maybe 10 to 15 years ago there was a debate within RC 20 and then everything subsided.

212
00:35:54.000 --> 00:36:13.000
No body wants to touch it again. So I dedicate the methodological part to the memory of new and also this paper was motivated by a Long overdue methodological update.

213
00:36:13.000 --> 00:36:23.000
Of what I call now a quasi. Right, rank, approach to doing discrete data analysis.

214
00:36:23.000 --> 00:36:29.000
I started that about 3 decades ago when I, soon after I joined RC. I was uncomfortable with it.

215
00:36:29.000 --> 00:36:45.000
Not satisfied with it until this paper. So this paper has a methodological thrust to it that is highly consequential to me, to my in memory of L Goodman and my affiliation with RC.

216
00:36:45.000 --> 00:36:58.000
Without further ado, let me start with the substantive Focus the substantive part of today's paper.

217
00:36:58.000 --> 00:37:06.000
This is joined work with my 2 former students, previous students, one pound and some gene.

218
00:37:06.000 --> 00:37:14.000
All right, what's the research question? We, this question is very boring. Seemingly boring.

219
00:37:14.000 --> 00:37:39.000
It's the simple classic question of the level and friend in in the generational mobility. That's the normal term and we chose to call in intergenerational status, because the measure we are going to use, the statistical measures that have been in place even in the Latina modern tradition has always been about closure associated

220
00:37:39.000 --> 00:37:44.000
dependency rather than mobility per se.

221
00:37:44.000 --> 00:38:02.000
So to us, IGC is the intergenerational class or occupational status association or some people call it dependency, which is the inverse of relative mobility or ferocity, net of structural mobility.

222
00:38:02.000 --> 00:38:12.000
That's the classical language in in RC. 28. The scope of today's supplementary topic is about 2 societies.

223
00:38:12.000 --> 00:38:21.000
The most advanced economy. The United States and the largest and fastest most dramatic developing country in human history.

224
00:38:21.000 --> 00:38:38.000
Yes China. But there are 2 methodological challenges behind it which I will close over. One is How do we do comparative intergenerational mobility research with discrete data?

225
00:38:38.000 --> 00:38:51.000
Without being biased by so-called marginal distribution differences. Contexts. The second challenge is the last differences.

226
00:38:51.000 --> 00:39:00.000
Across generations, parent generation and children in recent cohorts and societies.

227
00:39:00.000 --> 00:39:18.000
Now, there are a couple. Thanks to remember China Once upon a time, during the earliest stage of China research, was regarded as having lower IGC than the United States.

228
00:39:18.000 --> 00:39:30.000
That was done by up with a plow, a major player at that time. But that kind of finding what we mean a puzzle.

229
00:39:30.000 --> 00:39:38.000
People don't believe it. Even though it's done by one of the giants, thinkers at that time.

230
00:39:38.000 --> 00:39:50.000
Subsequent studies find that actually China As a whole, not urban China, but China as a whole has higher in that much higher RGC than the United States.

231
00:39:50.000 --> 00:40:01.000
Including the later study by one of our presenters today. The earlier paper by us song and SIR.

232
00:40:01.000 --> 00:40:05.000
Now.

233
00:40:05.000 --> 00:40:13.000
Let me repeat the methodological imperatives under the spotlight when we do this comparison is we must be able to account for.

234
00:40:13.000 --> 00:40:31.000
Margin difference marginal distribution differences between countries across generation and over cohorts. And we also hope to pay deal with back to the recently emphasized theme of multi dimensional nature.

235
00:40:31.000 --> 00:40:47.000
Of in the generational closure. There was also one of the theme in the second paper. Limitation of the existing approaches are There's the margin free imperative for discrete data.

236
00:40:47.000 --> 00:40:58.000
Bob Houser at Wisconsin school emphasize that a lot and highly embraced the government approach because of this beauty in resolving the margin III.

237
00:40:58.000 --> 00:41:12.000
Addressing the margin 3 imperative in categorical data analysis. However, we know today by a half century old mathematical theorem called Scalar's Them.

238
00:41:12.000 --> 00:41:20.000
This Plus, not true. Or contingency table approaches. Developed by Goodman and his associates.

239
00:41:20.000 --> 00:41:29.000
Cannot possibly be margin free, period.

240
00:41:29.000 --> 00:41:40.000
There is an alternative reason popular approach. Well, a Mount Economist, especially doing income analysis. They call it the Wangangangangang approach.

241
00:41:40.000 --> 00:41:44.000
However, sociologists try to adopt Bangladesh opposed to discrete data. I did that 30 years ago.

242
00:41:44.000 --> 00:42:05.000
I call it a quasi. Rain, approach. It failed to be modern free again. Even though the ideal rendering approach is indeed margin 3 according to scale of prayer as a special case.

243
00:42:05.000 --> 00:42:11.000
The quasi ramware pose is not margin 3.

244
00:42:11.000 --> 00:42:19.000
Adding to that problem of finding a modern-free approach with this great data, we have to take into account multi-dimensionality.

245
00:42:19.000 --> 00:42:33.000
Of course, not linear approach, including the latest condition, conditional logic model of Joe and share in AGS is exemplary of the state of the art in that paradigm.

246
00:42:33.000 --> 00:42:43.000
The alternative is the quasar membrane approach. impressed by song and says recent work.

247
00:42:43.000 --> 00:42:50.000
Okay, what is my alternative? I wouldn't say much except for this single slide. Tools rise for you.

248
00:42:50.000 --> 00:43:00.000
To save time, let me rapidly call the approach called LACOB approach. In 4, it's a shorthand for Leighton.

249
00:43:00.000 --> 00:43:06.000
My very,pler regression framework, lac cop.

250
00:43:06.000 --> 00:43:13.000
Layton is the P word. Cop stands for Coppola. That's another keyword.

251
00:43:13.000 --> 00:43:23.000
I'm using the theory of couplet theory in in Mathematics as well as financial statistics.

252
00:43:23.000 --> 00:43:43.000
A couple worth. Lateen variable approach to connect discrete observed data with an underlying continuous distribution. That is truly to which we can truly deal with margin free.

253
00:43:43.000 --> 00:43:48.000
Measurement of dependency.

254
00:43:48.000 --> 00:43:59.000
It's the second part of this approach is a graphical representation of the concept of ITC. I love this part.

255
00:43:59.000 --> 00:44:06.000
I would skip over all the mathematics because of time.

256
00:44:06.000 --> 00:44:14.000
This is part of the whole approach. I'm using a DAG diagram for the first time.

257
00:44:14.000 --> 00:44:29.000
To conceptualize IGC. As overall Oh, you still mouse? Yeah, this is for overall descriptive measure of IGC.

258
00:44:29.000 --> 00:44:44.000
This is the lack of approach to measuring IGCS. As dependency margin free dependence between 2 continuous latent variables.

259
00:44:44.000 --> 00:45:00.000
And then the other concept is called conditional IGC a not condition I mean sectoral IGC referring to the need to study components of overall IGC.

260
00:45:00.000 --> 00:45:18.000
Sectoral in this case is going to mean Farm to Nump farm transitions. Dependency of farm non farm status from parents The father, 2 children.

261
00:45:18.000 --> 00:45:30.000
It can be represented in day format, a causal diagram. And then there's something called non sectoral IGC, which is the residual.

262
00:45:30.000 --> 00:45:40.000
Dependency after conditioning on sectoral relationships. Here. Observe Sacharo ITC.

263
00:45:40.000 --> 00:45:54.000
All this concepts are unfamiliar to people who used to do like linear modeling. But these are conventional concepts in modern-day counter-factual causal analysis.

264
00:45:54.000 --> 00:46:07.000
And I'm merging the 2 tradition. Attribute to the application. Oh, these are the details I won't go into.

265
00:46:07.000 --> 00:46:18.000
I won't go over the data. I will go to the finding for you. The first finding enabled by this approach.

266
00:46:18.000 --> 00:46:32.000
Despite the beautiful comparison. Between China and Western countries expressing the United States in recent work. We are missing something.

267
00:46:32.000 --> 00:46:45.000
Do we have a racial scale metric easily interpretable? That can be made that allow us to compare RGC between United States and China.

268
00:46:45.000 --> 00:46:56.000
Our answer is yes. There's a simple measure called Kendose Tower. Kendall's Tower, which is a non-parametric measure of Margin Free Association.

269
00:46:56.000 --> 00:47:11.000
It varies from minus one to positive one with racial scale meaning there's an absolute 0 there. So the number you see here are not arbitrary.

270
00:47:11.000 --> 00:47:27.000
Do United States is on that scale just about less than point 2. Of IGC. In contrast, China started out for the older first cohort start out as point 4 over 2 times.

271
00:47:27.000 --> 00:47:36.000
The IGC, they experienced, compared to people in the United States. And that trend kept declining.

272
00:47:36.000 --> 00:47:45.000
So. Just from this diagram, you can see major comparison. Stability for the United States, even there were talks.

273
00:47:45.000 --> 00:47:55.000
There have been talks about rising. In mobility in the United States? Actually, if there's any rise, it's trivial.

274
00:47:55.000 --> 00:48:07.000
For China, there's massive decline of closure. That means massive increase of mobility. That is about overall trend.

275
00:48:07.000 --> 00:48:22.000
Secondly, I talk about sectoral in the generational closure and that was one of major theme in Joe and says AGS article on China.

276
00:48:22.000 --> 00:48:33.000
I leave what you saw earlier in the background in light, gray color and the new finding on the same scale of Kendo's town.

277
00:48:33.000 --> 00:48:45.000
In this way. The United States has by point estimate At least as high and even higher.

278
00:48:45.000 --> 00:48:56.000
Sectoral closure that is farmland farm area then farmers in China.

279
00:48:56.000 --> 00:49:06.000
They may not be a certificate due to the sparse data in the US case but the estimate for China is very sharp.

280
00:49:06.000 --> 00:49:18.000
Very small confidence interval. This is our second finding. It corroborates our intuition that China has massive sectoral area.

281
00:49:18.000 --> 00:49:27.000
That was one of the main messages of Joe and Share. Further down the road. This is unique to us.

282
00:49:27.000 --> 00:49:37.000
We can assess. After necking out sectoral burial, what is the degree of IGC in China?

283
00:49:37.000 --> 00:49:47.000
10 min. Hands up. Oh, sorry. Let me hurry up.

284
00:49:47.000 --> 00:50:01.000
So non-sectoral IGC after necking out sectoral. This is the picture. Both US and China are on par.

285
00:50:01.000 --> 00:50:13.000
Can you imagine that? In a margin treatment by a margin free measure and the racial scale mesh of IGC, they are the same.

286
00:50:13.000 --> 00:50:22.000
And that has been true. For all those cohorts and this even decline in recent coal. The latest score.

287
00:50:22.000 --> 00:50:32.000
This is a dramatic finding, will not be possible nor plausible, I guess, without a consistent framework that address overall, sectoral and non sectoral.

288
00:50:32.000 --> 00:50:50.000
I can also decompose the contribution and tell stories about it. If I may, let me jump to The final conclusion.

289
00:50:50.000 --> 00:50:59.000
Paradox. Net of farm non-fung barrier. The American and Chinese have faced almost identical level.

290
00:50:59.000 --> 00:51:10.000
And persistence of non sectoral IGC. For 4 decades. Of Perhaps. Put it in layman term, a lock slide.

291
00:51:10.000 --> 00:51:25.000
Roughly speaking. Urban Chinese under Communist regime. Despite all the changes in institutions and so on. They have been living alive as

292
00:51:25.000 --> 00:51:36.000
Close to an American dream. As Americans. In post war period. Is a paradoxical finding.

293
00:51:36.000 --> 00:51:42.000
And it's true. I've done lots of robust and shit and it's true. It's very robust.

294
00:51:42.000 --> 00:51:55.000
I'm leaving at that there are other potentials on the methodological side as I said we have about 10 papers in the working at different stages of development.

295
00:51:55.000 --> 00:52:23.000
Or taking different tags to extend their basic approach here and I welcome your comments. And suggestions. Thank you.

296
00:52:23.000 --> 00:52:36.000
Okay, oh. Staff here. So we have the force paper stamp in Brazil.

297
00:52:36.000 --> 00:52:39.000
Alright.

298
00:52:39.000 --> 00:52:46.000
Good morning to our field there in Australia and Brazil here at the evening. And thank you, professors.

299
00:52:46.000 --> 00:52:49.000
Are you listening, Shane?

300
00:52:49.000 --> 00:52:50.000
Yes.

301
00:52:50.000 --> 00:53:08.000
Yeah, okay. Thank you. So thanks to everyone attaining this session. It's an honor to be here and I will present, my research with I mean they quality pay and occupation amongst 10 graduates in Brazil.

302
00:53:08.000 --> 00:53:29.000
So first, a very brief context. Often in part in Brazil. Brazil is a very Nico society and despite some important advances In acquired reductions during the first decade of this millennium, we remain with Go intergenerational mobility and with high income inequality and comparing shorter countries that provide information to

303
00:53:29.000 --> 00:53:32.000
OCD.

304
00:53:32.000 --> 00:53:42.000
So just some data contest for Data here. We can see that in Brazil it takes on average 9 generations.

305
00:53:42.000 --> 00:53:49.000
For both born and low income families to oppose an income. Why are they OECD average is half of that?

306
00:53:49.000 --> 00:53:59.000
4.5 generations. And when you are also from a WCD. We can see that Brazil has the highest wage premium for completing higher education.

307
00:53:59.000 --> 00:54:07.000
Remember Zoom those that can be higher in location have an average 4 times the income of those that have only high school.

308
00:54:07.000 --> 00:54:14.000
Why are the OECD average is also half of that 2 times.

309
00:54:14.000 --> 00:54:25.000
So back to our research question. Give this contacts, we stand degrees, contribute to inter-generational mobility in Brazil.

310
00:54:25.000 --> 00:54:38.000
First, let us explain why we focus almost 10 degrees. Firstly, for TORRECT provisions, some research suggests that these degrees are relevant to your videos to get better paid and more procedures jobs.

311
00:54:38.000 --> 00:54:48.000
It would also favor inter-generational mobility because technical skills is to get the job done are more likely to be acquired through education.

312
00:54:48.000 --> 00:54:56.000
And secondly, 4 practical reasons. That is data availability. Is 10 graduates of 2,011.

313
00:54:56.000 --> 00:55:05.000
Foot a socioeconomic questioner as part of a nationwide evaluation test of their programs.

314
00:55:05.000 --> 00:55:16.000
So let's take a look at our data. And we joined several administrative data sets. What see sometimes short, all focus only in 2 of them.

315
00:55:16.000 --> 00:55:35.000
One from the location of side. Feel it by higher education institutions and by students. And the second from the labor market side, These join data sets give us information on 62,000 individuals that's graduated from stem fields in 2,011.

316
00:55:35.000 --> 00:55:48.000
Which at the time were between 18 and 40 years old. We also have information informal employment of Disney videos between 2,007 and 2,018 that is.

317
00:55:48.000 --> 00:56:06.000
From 4 years before they graduated to 7 years after the graduation. Use that allows us to analyze how they're occupation if they change it over time and thus to portray the next generation of mobility.

318
00:56:06.000 --> 00:56:14.000
We also conducted logging in your MOCC level analysis. To estimate the influence of higher education on their intergenerational mobility.

319
00:56:14.000 --> 00:56:28.000
For the most level models, we had 40,000 individuals with complete socio economic information. That also were employed in 2 different periods after graduation.

320
00:56:28.000 --> 00:56:39.000
So this chart display the 16 stem fields that we consider in this. In the yellow bars here at the top.

321
00:56:39.000 --> 00:56:50.000
We can see the first, the first generation students and they blue bars. Below we see the second generations students.

322
00:56:50.000 --> 00:56:56.000
Sorry, that's the hopeless concern. The blue bars the first generation is and the yellow bar is the second generations.

323
00:56:56.000 --> 00:57:13.000
But for generation we mean students whose both parents have not completed higher education. We also show graduates average unit earnings considering selling years after the graduation on the red line.

324
00:57:13.000 --> 00:57:26.000
These 10 fields are also ordered by earnings. We can see that on the left side. Yeah, and there is a concentration of fuse in which more than 7% of the graduates are from first generation.

325
00:57:26.000 --> 00:57:36.000
This feuds are science and math related. In addition to information technology, associate degree. But also just next to them.

326
00:57:36.000 --> 00:57:50.000
Is the few which has the highest proportion of students from second generation architecture. With 62%. Of students from second generation, but with average earnings similar to physics.

327
00:57:50.000 --> 00:58:01.000
Which have only 29% of the students from second generation. The highest earnings are from the traditional engineering fields.

328
00:58:01.000 --> 00:58:13.000
On the right, which had around half of these students from first generation. On the whole sample, we have an average of 63% of students of first generation.

329
00:58:13.000 --> 00:58:25.000
The fact that we have mostly first-generation graduates reflects the expansion of higher location in Brazil during the 19 nineties and the 2 thousands.

330
00:58:25.000 --> 00:58:35.000
Now these 4 shorts allows us to track inter-generation will be for both first and second generation graduates.

331
00:58:35.000 --> 00:58:49.000
First generation is here on the left. I'm second generation graduates here on the right. We can see that most of first generation graduates were already working before graduation.

332
00:58:49.000 --> 00:59:01.000
67% of them were formally employed. The same is not true for second generation. Only 40% of them We're formally employed before graduation.

333
00:59:01.000 --> 00:59:09.000
They graduated in 2,011. So the first bar portrays, their employment situation before that waving.

334
00:59:09.000 --> 00:59:21.000
I compound the first and the second bars. We see how their employment situation changes after graduation. The overall train is of upward mobility.

335
00:59:21.000 --> 00:59:29.000
We can see how the number of unemployment decrease in the number of graduates occupying professions and managerial positions.

336
00:59:29.000 --> 00:59:38.000
45 as EGP one and 2 increases. What's the point for mobility is higher for certain generation Broadways.

337
00:59:38.000 --> 00:59:55.000
Who are more likely to reach professional in managerial quotations. Downward mobility is very low, indicating that the initial labor market situation plays a relevant role in future trajectory.

338
00:59:55.000 --> 01:00:10.000
This other shot shows the show that Ernie's house increase over time, especially after graduation. The army's level is strongly associated with gender parental education and race.

339
01:00:10.000 --> 01:00:21.000
But with gender above all. Men have higher average earnings than women. Despite of your social origin or race.

340
01:00:21.000 --> 01:00:36.000
Good change of earnings after graduation, however, is more associated with parental education. Which second generation graduates having higher percentage increases in their yearnings after their graduates?

341
01:00:36.000 --> 01:00:55.000
Now we jump to some of the results of the Mood. This circles here. Has shown the total effects of gender race and generation why the triangles showed in the direct effect mediated by characteristics of higher education institutions and programs.

342
01:00:55.000 --> 01:01:06.000
Several characteristics of a higher education institutions and programs were considered using the models, including this their selectivity, the percentage of professors with a PhD.

343
01:01:06.000 --> 01:01:18.000
If you pick up public or private among other We see that that the generation effects is almost completely mediated.

344
01:01:18.000 --> 01:01:33.000
By higher education, which means that when we take into account characteristic of institutions and programs, the completion of higher education, similar terms for first and second generation graduates.

345
01:01:33.000 --> 01:01:46.000
This indicates that current communication difference in learning is observer regularly in the descriptive statistics are largely explained by the access to different pockets of higher education.

346
01:01:46.000 --> 01:01:54.000
The same is not true, however, for race and especially for gender, which are only partially mediated by higher education.

347
01:01:54.000 --> 01:02:06.000
Generating a quiet remain very high. And also increase over time. You hear the gray bars is for the first period between 2,012 and 2,014.

348
01:02:06.000 --> 01:02:18.000
And then black bars, is for the second period here between September 16, 2,011, 2,018, sorry.

349
01:02:18.000 --> 01:02:32.000
This last chart shows only parental education effects, but for groups of gender and race. For plotting this chart, we basically run models with international effects for gender, parental education and race.

350
01:02:32.000 --> 01:02:45.000
You change the reference category of interactions term. You mean line of the chart. Then we can see that the mediation effect is complete for men.

351
01:02:45.000 --> 01:02:52.000
Yeah. Independently of race, but not for women.

352
01:02:52.000 --> 01:03:00.000
That is when we adjust by higher education quality, first and second generation men have similar earnings, but first generation women.

353
01:03:00.000 --> 01:03:12.000
Unless then the second generation counterparts. Showing that there is an important interaction effects between gender and parental education.

354
01:03:12.000 --> 01:03:25.000
Now the final remarks here. Occupational flows review that graduation proves employment equality How.

355
01:03:25.000 --> 01:03:37.000
Broadway displays great upward mobility. Low-oward mobility indicates that occupations immediately after graduation plays a crucial role in future opportunities.

356
01:03:37.000 --> 01:03:44.000
Early zoom equality due to any video character issues were substantially mediated by quiet of higher education for men.

357
01:03:44.000 --> 01:03:48.000
The fact of parental

358
01:03:48.000 --> 01:03:57.000
Oh, for me, the effect of current current location was completely mediated by higher education. However, gender and education was completely mediated by higher education.

359
01:03:57.000 --> 01:04:00.000
However, gender and education was completely mediated by higher education. However, gender and race in aquatics were only partially mediated.

360
01:04:00.000 --> 01:04:13.000
Generating acquired remain very high. The Spain losing effect over time. This confirms previous research showing a pattern of discrimination against women in the labor markets.

361
01:04:13.000 --> 01:04:22.000
Meation effects review that most parental education inequalities are due to its proportional location of first generation videos.

362
01:04:22.000 --> 01:04:34.000
Into less prestigious programs. However, when using videos to successful fully complete higher education The earnings are similar to those individuals coming from more favorable backgrounds.

363
01:04:34.000 --> 01:04:46.000
This shows that investing quality of education and ensuring better access and completion rates for this but disadvantage groups and possible to reduce income inequalities in Brazil.

364
01:04:46.000 --> 01:04:58.000
That's it.

365
01:04:58.000 --> 01:05:09.000
Okay, thank you. So we have this last presentation.

366
01:05:09.000 --> 01:05:17.000
It's a with you. I'm just gonna try. No.

367
01:05:17.000 --> 01:05:29.000
So, you see?

368
01:05:29.000 --> 01:05:42.000
Our presenter which is remote Is he here? Wait, please share your screen if

369
01:05:42.000 --> 01:05:48.000
You are here.

370
01:05:48.000 --> 01:05:54.000
He's not here. He's not here.

371
01:05:54.000 --> 01:05:55.000
Smoke.

372
01:05:55.000 --> 01:05:58.000
Hi.

373
01:05:58.000 --> 01:05:59.000
Should we?

374
01:05:59.000 --> 01:06:03.000
So is there are a video or no video?

375
01:06:03.000 --> 01:06:11.000
I don't think there is a video cause we just tried this and it doesn't play so

376
01:06:11.000 --> 01:06:12.000
Okay.

377
01:06:12.000 --> 01:06:20.000
It seems like he's just gonna skip it. We probably should get started discussion.

378
01:06:20.000 --> 01:06:24.000
Do you want to say if you would? Professor.

379
01:06:24.000 --> 01:06:34.000
So probably wish we can just, open the floor for QA. We still have plenty of time, right?

380
01:06:34.000 --> 01:06:39.000
Let's see. Are

381
01:06:39.000 --> 01:06:47.000
I'm a bit confused about the time difference. So now is 9 41 and the session and team 1030, right?

382
01:06:47.000 --> 01:06:49.000
Wendy.

383
01:06:49.000 --> 01:06:56.000
1020, okay. So we still have a about 40 min for QA.

384
01:06:56.000 --> 01:07:09.000
There is a online question here for Oh Okay, a for Professor Mortimer. How do you control?

385
01:07:09.000 --> 01:07:13.000
For sibling size.

386
01:07:13.000 --> 01:07:18.000
Oh, you are muted. You muted.

387
01:07:18.000 --> 01:07:36.000
No, we didn't control for siblings. Or family size. We only, control for family structure comparing single parent families and dual parent families.

388
01:07:36.000 --> 01:07:39.000
Okay, thanks.

389
01:07:39.000 --> 01:07:55.000
And we have some other questions from our Audience here.

390
01:07:55.000 --> 01:08:11.000
Well, if no one ask a question, so I also have a question about the study. I wonder whether some measures also our self direction was just work related values.

391
01:08:11.000 --> 01:08:21.000
May also change over the lifetime and whether that will change the estimates of the inter-generational correlation or transmission of these values.

392
01:08:21.000 --> 01:08:33.000
Yeah, this is something that we could assess in. Future research as we have, you know, information that I presented was, in the mid thirties.

393
01:08:33.000 --> 01:08:34.000
The party is right.

394
01:08:34.000 --> 01:08:46.000
It goes on until, the mid forties. And so we could. Assess that but we haven't yet done so

395
01:08:46.000 --> 01:08:55.000
Or if you just have multiple measures, can you see whether there is some lifetime fluctuations or even a trend?

396
01:08:55.000 --> 01:09:07.000
I think in some of the earlier status attendant models by Houser and others, they typically have multiple measures but are assume all the fluctuations are cost by measurement errors.

397
01:09:07.000 --> 01:09:18.000
But it's also possible that children started to develop those values as they age. So you may see a weaker correlation intergenerational correlation.

398
01:09:18.000 --> 01:09:23.000
Between parents and children in their say twenties if they have to measure and the correlation makeup becomes stronger as you measure children's values in their fortys.

399
01:09:23.000 --> 01:09:33.000
So it will be interesting to see if there is this live course pattern.

400
01:09:33.000 --> 01:09:49.000
This could be, in the paper, you know, we point out that, especially in the United States, there's a lot of, floundering around and, not settling into an occupation, you know, until fairly late.

401
01:09:49.000 --> 01:10:05.000
I mean, unlike systems that have clear school to work transitions and you know the dual system in Germany with you know strong vocational education and so on and in the United States there's much more you know shifting and movement between occupations.

402
01:10:05.000 --> 01:10:34.000
So that's why we decided that instead of looking at you know, occupational. Choices or you know destinations or values you know in the early you know in the in the late teens or in the early twentys that we would wait until careers had become established, but this is something that we can look at in the

403
01:10:34.000 --> 01:10:41.000
future, especially moving forward to.

404
01:10:41.000 --> 01:10:45.000
No, thank you for that comment.

405
01:10:45.000 --> 01:10:53.000
So currently you only use the data until 2,011, right? But the data you said are collected onto 2,019.

406
01:10:53.000 --> 01:10:54.000
That's right.

407
01:10:54.000 --> 01:10:59.000
So, nearest the wave have not been made of publicly available.

408
01:10:59.000 --> 01:11:12.000
Right, the most recent, data have been deposited in the ICPSR system, but they haven't made it available yet because they do a lot of checking and it's been sitting there for quite a long time.

409
01:11:12.000 --> 01:11:20.000
But if anyone's interested in looking at the data, I'd be happy to share, information through our, system.

410
01:11:20.000 --> 01:11:27.000
Okay. Thank you.

411
01:11:27.000 --> 01:11:31.000
I have a question for, Louise.

412
01:11:31.000 --> 01:11:46.000
So, very interesting study because I myself is also interested in stamp. I was wondering if you compared your mobility chains in stand with other fields such as social science and humanities.

413
01:11:46.000 --> 01:11:59.000
Because we know that and promote social mobility to some extent. But since you are focusing on gender and there is really an eco-gender distribution in stem non stem.

414
01:11:59.000 --> 01:12:06.000
So we don't have a reference. As for how large this effect size is in stem compare with other fields.

415
01:12:06.000 --> 01:12:10.000
So what do you think?

416
01:12:10.000 --> 01:12:12.000
Oh, you're muted by the way.

417
01:12:12.000 --> 01:12:28.000
Yeah, yeah, thank you very much for the question. Yeah, I think that makes total sense. That makes a lot of sense to do this comparison between Stan and other fields, but our data here was only about 10, but so we didn't do this comparison, but we plan to.

418
01:12:28.000 --> 01:12:40.000
We intend to do this. Not in future works. And I, yeah. I expect that Under fields maybe we haven't.

419
01:12:40.000 --> 01:12:46.000
Not so much mobility as a 10 fields, but let's see.

420
01:12:46.000 --> 01:12:52.000
Oh, thank you.

421
01:12:52.000 --> 01:13:05.000
Oh, we have, Tony here. In person question. Since, several of the papers used with the role of skill.

422
01:13:05.000 --> 01:13:22.000
In stratification of like to call your attention to the following observation. Number one in as in my own presentation having a large effect.

423
01:13:22.000 --> 01:13:31.000
A Savian effect. By statistical standard doesn't mean it's important, substantively important.

424
01:13:31.000 --> 01:13:45.000
But just like in the US case Sataro Barrier is even stronger, right? Some measure. Dan, Sectoral buried that is found in Bambara than in China.

425
01:13:45.000 --> 01:13:55.000
But sectoral parent is extremely consequential in interpreting both the level and the dramatic declining trend.

426
01:13:55.000 --> 01:14:05.000
Of IGC in main in China. Absolutely no contribution. Shouldn't tap the stupidity of trap.

427
01:14:05.000 --> 01:14:13.000
In the US case. So when we start using utilizing skills data, which I did 10 years, started 10 years ago as well.

428
01:14:13.000 --> 01:14:25.000
I was keep wondering where to put my money. Having lots of measure of skill where what can it really explain?

429
01:14:25.000 --> 01:14:39.000
I saw quite a few attempts to introduce skill over hunters and us paper and others. Mastering skill and then projecting how consequential it is.

430
01:14:39.000 --> 01:14:48.000
If we ignore them, we miss certain big picture. How about in the context of the new economy? Economists are doing that a lot.

431
01:14:48.000 --> 01:15:04.000
We sociologists are trying to build in this component. Where will skill? Help us understand the new stratification structure, say in the United States.

432
01:15:04.000 --> 01:15:13.000
The background of my person, another pattern of my question is in finding Reporter I think is on Monday.

433
01:15:13.000 --> 01:15:25.000
Hi, my collaborator, Oneong Pong, in a comparison between China and US. New. Binary cost structure.

434
01:15:25.000 --> 01:15:39.000
The So. Kind of the vanishing barrier between white color, routine white college occupations.

435
01:15:39.000 --> 01:15:58.000
With 10 manual jobs. Blue collar and white color boundaries are shipped have totally shifted very clearly And that's based on the same methodology that I use for my China US comparison a moment ago.

436
01:15:58.000 --> 01:16:06.000
His finding would impact he can speak for himself a little later would imply that Wow, all the new berries.

437
01:16:06.000 --> 01:16:27.000
Are concentrated between the so-called professional managerial class Up there versus everybody else. How well can Skew data the management of skill of different occupations and people's job can help us understand this new barrier.

438
01:16:27.000 --> 01:16:36.000
Emergent buried within the American car structure. So that that is a long way to say that I see a lot of promise.

439
01:16:36.000 --> 01:16:45.000
Introducing skill. I just want to. Ask your insight in how do you The while sense of significance.

440
01:16:45.000 --> 01:16:56.000
In a quantitative way see if you can apply in a quantitative way Thank you.

441
01:16:56.000 --> 01:17:15.000
Well, so are. Regarding skills, we believe a lot of the intergenerational transmission It's not only about occupational categories, say parents, maybe doctors and children may be nurses.

442
01:17:15.000 --> 01:17:28.000
So they are in different occupational categories but they are still connected within their industry or other kind of skills or you say parents is a surgeon whereas say children is a craftsman.

443
01:17:28.000 --> 01:17:37.000
So they may be in very different occupations, but still they are connected with say manual skills or dexterity.

444
01:17:37.000 --> 01:17:54.000
So that means in many ways, well, if we just a focus on the diagonal sales in our our mobility table we may overlook inter-generational associations are in those off diagonal cells.

445
01:17:54.000 --> 01:18:18.000
That are connected by users say work values or job tasks or skills or work content or work environment. So in our study we did not only measure skills but we also measured or this are several dozens of this dimensions and eventually we reduced them to our 11 latent factors.

446
01:18:18.000 --> 01:18:31.000
So that's a motivation of our paper. But I also are want to hear your thoughts about why this are you said cannot explain.

447
01:18:31.000 --> 01:18:38.000
So changing our Association or can elaborate on that?

448
01:18:38.000 --> 01:18:54.000
Yes. The best. Wait to explain it is you use model comparisons to indicate whether introducing skill and related information to improve model fit.

449
01:18:54.000 --> 01:19:13.000
That is the classical way of 3 comparison and Justifying so-called at a new dimension to the explanation then it and breaches Goodness of people, that But it kind of hides behind.

450
01:19:13.000 --> 01:19:24.000
The sense of quantitative contribution. Just like in in my analysis I can tell My audience, my readers.

451
01:19:24.000 --> 01:19:35.000
How much does sectoral barrier count? Toward the overall contribute over all IGC in the country and it changes over time.

452
01:19:35.000 --> 01:19:47.000
I can precisely showed it on the same metric. Not just saying that it it is consequential. I can also show it quantitative how much it counts on account.

453
01:19:47.000 --> 01:19:55.000
So I'm looking forward to say in your paper. You can provide that sort of quantification.

454
01:19:55.000 --> 01:20:10.000
Oh, I see. So you want to know say based on say say, a bounded, say, proportion from 0 to one, how much can be explained by skills that I cannot be explained by, say, occupations.

455
01:20:10.000 --> 01:20:11.000
Yes, one.com. Yes, one.

456
01:20:11.000 --> 01:20:31.000
How much is? Like what you measure is based on say Kendall's Tower so we know that it's so value, found it between say 0 1 one so something about say point 2.4 would be considered as a significant level of association.

457
01:20:31.000 --> 01:20:32.000
Yeah.

458
01:20:32.000 --> 01:20:44.000
Right. Okay. We haven't done that, but I think that's a great suggestion. And that it can be very easily incorporated into our analysis, but of which means that we will know how to use learning your models.

459
01:20:44.000 --> 01:21:08.000
So we were just a measuring her generation associations based on the continuous measures of are skills rather than are incorporated into our local new model as Mike, did in his classic paper based on this SAT model as Mike, out in his classic paper based on this SAT model that also incorporated our skills.

460
01:21:08.000 --> 01:21:20.000
Our end tasks into inter-generational correlations of occupations. Are, but I like such suggestion and are, we will add it to our paper.

461
01:21:20.000 --> 01:21:24.000
Thank you.

462
01:21:24.000 --> 01:21:38.000
Just a clarification question. Did I hear it correctly that the conclusion is that when one wants to inherit their parents occupational skills.

463
01:21:38.000 --> 01:21:50.000
Then they have to make vertical sacrifices. Is that one of your findings? It is a little bit any intuitive if I can inherit my parent skills.

464
01:21:50.000 --> 01:22:00.000
It would help me make vertical improvement on my status, right? Why would it be? Yeah, I have to like.

465
01:22:00.000 --> 01:22:02.000
No mobility.

466
01:22:02.000 --> 01:22:15.000
I'm not sure we have that finding. We show that for some of skills serious. Our positive correlation, some have negative correlation because it's also varied by gender.

467
01:22:15.000 --> 01:22:29.000
Say so far there is a manual worker with very strong say our manual skills so the daughter may want to avoid our being say a manufacturing worker but instead become a service worker.

468
01:22:29.000 --> 01:22:36.000
So, are, so the inter-generational correlation sometimes can be positive, sometimes can be negative.

469
01:22:36.000 --> 01:23:00.000
It's not always positive. As typically in intergenerational occupational studies, that typically assumes a correlation is positive and there is a rare discussion about negative correlations but this is more common among our parent and offspring of different or sex like mother sons or father daughters.

470
01:23:00.000 --> 01:23:03.000
Got it. Thanks.

471
01:23:03.000 --> 01:23:10.000
Other questions? Other questions?

472
01:23:10.000 --> 01:23:14.000
Oh yes, please come ahead.

473
01:23:14.000 --> 01:23:20.000
Based on, for the in town's paper. I'm just curious about it.

474
01:23:20.000 --> 01:23:29.000
So is it okay to? And all the relationship between Orleans and the destination across the derivatives of education.

475
01:23:29.000 --> 01:23:41.000
Beauty when you are mess with it. I don't know, our father application we are messed up to investigate that for example OED Association.

476
01:23:41.000 --> 01:23:55.000
And is it okay to decompose the association into? Some part, like, cross-imandices.

477
01:23:55.000 --> 01:24:03.000
Yeah, so the final screening that I see if we can

478
01:24:03.000 --> 01:24:12.000
Final. That's true. Let's see. The answer is yes.

479
01:24:12.000 --> 01:24:25.000
It's one of the half dozen extensions resulting in about 10 papers.

480
01:24:25.000 --> 01:24:37.000
Purpose of the great utility of conceptualizing IGC in a path diagram is precisely because we intended to inform causal analysis.

481
01:24:37.000 --> 01:24:43.000
We can do conditioning and so on, introducing additional dimensions and control mediating factors and so on.

482
01:24:43.000 --> 01:24:52.000
In that context, very consistently and explicitly.

483
01:24:52.000 --> 01:24:59.000
Here I we can extend.

484
01:24:59.000 --> 01:25:19.000
The analysis from ITC in occupational class status to education. And also we lay and other kind of intergenerational closure such as wealth.

485
01:25:19.000 --> 01:25:25.000
You name it, all the classical things we have talked about in the linear monitoring context can be transferred to this.

486
01:25:25.000 --> 01:25:32.000
And do it in a truly margin-free manner. There's a first part of the answer.

487
01:25:32.000 --> 01:25:39.000
The second part is. We can also relay across different dimensions. Education, I just like my question yesterday at this room.

488
01:25:39.000 --> 01:25:56.000
Educational ITC has a tremendously important role to play and we can quantify its contribution to occupational or status or class IGC.

489
01:25:56.000 --> 01:26:10.000
And we can also specify how much does this kind of intergenerational ideas in education can be distinguished from the mediating effect of children's education.

490
01:26:10.000 --> 01:26:22.000
All this because of the DAG representation. It's beautiful. We haven't done that in a lot in a model in contact because we focus so much on pure association.

491
01:26:22.000 --> 01:26:35.000
But once you translate in a DAG conceptualization, then everything becomes sensible, interpretable and provide good guidance on how to pursue what to do and what not to do.

492
01:26:35.000 --> 01:26:50.000
So that's my lines.

493
01:26:50.000 --> 01:26:52.000
Okay, we still have a few minutes.

494
01:26:52.000 --> 01:27:11.000
Yeah, well, so I have some questions for our Tony, our presentation. I wonder if you, well, so first of all, I think so finding are is very provocative.

495
01:27:11.000 --> 01:27:12.000
Yeah.

496
01:27:12.000 --> 01:27:23.000
It's our or so solid provoking or It seems very different from our findings. Are and also some of the media portrait of the mobility trend in contemporary China.

497
01:27:23.000 --> 01:27:34.000
By the way, it seems to me the correlation based on, It's a bit low particularly for the US.

498
01:27:34.000 --> 01:27:39.000
So your estimates is about point 2 whereas in. Most are mobility studies at least for the US.

499
01:27:39.000 --> 01:27:54.000
So the inter-generational correlation or IG is about point 4 or even based on Rajadi some more recent studies point 5 or even higher.

500
01:27:54.000 --> 01:28:12.000
We asked for China, we typically sync are the inter-generational correlation is smaller.

501
01:28:12.000 --> 01:28:27.000
Than those estimates in the literature. Is it because you use the ISE rather than say earnings or our occupational scores.

502
01:28:27.000 --> 01:28:38.000
Okay, again that 2 passed to the answer we thought hard about it the first thing is I GE is definitely not margin free.

503
01:28:38.000 --> 01:28:43.000
Awesome, Brennan opposed, that is.

504
01:28:43.000 --> 01:28:53.000
Using an analogy of the income bank correlation or regression applying it to Discrete data.

505
01:28:53.000 --> 01:29:10.000
That's not. The margin free property. Of income rank, which is continuous. You have really uniformly distribute user unique area for each individual case in the observation.

506
01:29:10.000 --> 01:29:17.000
That's not true when we have discrete data, occupational data, occupational categories and so on.

507
01:29:17.000 --> 01:29:26.000
So that is one complication to it. The second one is income is of course different from class. Or occupational status group.

508
01:29:26.000 --> 01:29:38.000
So if the level differs between Chatty's finding and our finding this Financial. The question is where the the difference comes from.

509
01:29:38.000 --> 01:29:49.000
That is a separate empirical question to answer. It's not a methodological question. This is part one of my answer.

510
01:29:49.000 --> 01:30:03.000
So. The second part of my answer is your finding and recent findings by by by recent work on American rise in inequality.

511
01:30:03.000 --> 01:30:13.000
Our answer is There are 2 components where if you look at the Status, occupational class or status.

512
01:30:13.000 --> 01:30:28.000
Hi GC Trent is remarkably stable if there's any change in the Larger scheme of things from 0 to one is just Tribune over time.

513
01:30:28.000 --> 01:30:39.000
Secondly, I have to say, Johan says, recent work about rice is based on a method.

514
01:30:39.000 --> 01:31:00.000
What we call, Simple truncation. I have another paper. After this one and our paper go into detail, the paradox, the puzzling difference between UCS work and a recent work and hours.

515
01:31:00.000 --> 01:31:09.000
He projects a rise in inequality. We can identify the source of that rice. As being misleading.

516
01:31:09.000 --> 01:31:19.000
Is due to simple truncation. If you take the conditioning approach as you will do in the causal modeling framework.

517
01:31:19.000 --> 01:31:24.000
Is spread. You can reproduce its rice as well as eliminate the rice by changing the method of analysis.

518
01:31:24.000 --> 01:31:39.000
So the short answer. To the so called differences. We saw it really methodological. And we can independently play resolve that person.

519
01:31:39.000 --> 01:31:42.000
Thank you.

520
01:31:42.000 --> 01:31:45.000
I just want to follow up on this. I'm still a bit of puzzle, but I think I may need to study your message better.

521
01:31:45.000 --> 01:32:02.000
But typically when you exclude mobility cost by changing margins or the structural mobility, you will see even higher intergenerational correlation rather than lower.

522
01:32:02.000 --> 01:32:15.000
But for your messenger, once you create this margin free measures, you see more mobility rather than less mobility.

523
01:32:15.000 --> 01:32:16.000
But Addie Stewart.

524
01:32:16.000 --> 01:32:19.000
I did not show any comparison between the 2. Two-week property will be absolute mobility.

525
01:32:19.000 --> 01:32:32.000
You have to produce a measure of absolute mobility and then the marginally measure of IGC. I did not show the so-called word and without structural mobility.

526
01:32:32.000 --> 01:32:51.000
I did not show that. Or the IGC measures are margin free. Traditional the ideal of that I do target of where relative mobility is now transferred into the concept of ITC here.

527
01:32:51.000 --> 01:32:52.000
In our payment.

528
01:32:52.000 --> 01:33:01.000
I think you imply that if we compare your estimates with others that estimates You think others estimates are not a margin free.

529
01:33:01.000 --> 01:33:10.000
And they show. Even our higher intergenerational mobility. Actually based on your method, the margin free.

530
01:33:10.000 --> 01:33:14.000
Estimate should be higher than previous estimates.

531
01:33:14.000 --> 01:33:28.000
Oh, others at estimates. Aim to be supposedly margin 3 supposedly relative mobility we are comparing our measure with those supposed.

532
01:33:28.000 --> 01:33:43.000
Measures of relative mobility not compared with absolute mobility. All the linear modeling approach. Crane to, hope aspire to produce margin 3 measures.

533
01:33:43.000 --> 01:34:01.000
That is relative mobility rather than structural class relative. No. That was never to go. So we are comparing only worth the so-called quasi relative mobility you may see.

534
01:34:01.000 --> 01:34:10.000
It's not purely, but it's contaminated. It's not saying that it deliberately combine.

535
01:34:10.000 --> 01:34:16.000
By design combines both relative and structural mobility. No? We don't interpret that one.

536
01:34:16.000 --> 01:34:23.000
Okay, well so I think I need to study some misses better too understand how you decompose it.

537
01:34:23.000 --> 01:34:29.000
Well, yeah, so I will send an email to take a look at your slides. Or if you have a working paper.

538
01:34:29.000 --> 01:34:40.000
Thank you. Yep, thank you. We do have a working paper. We are about to finalize it.

539
01:34:40.000 --> 01:34:41.000
That's great.

540
01:34:41.000 --> 01:34:42.000
Thank you.