PreprintPDF Available

In And Out Of Unemployment -Labour Market Dynamics And The Role Of Testosterone IN AND OUT OF UNEMPLOYMENT -LABOUR MARKET DY- NAMICS AND THE ROLE OF TESTOSTERONE

Authors:
Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Biological processes have provided new insights into diverging labour market trajectories. In this paper, we use population variation in testosterone levels to explain transition probabilities into and out of unemployment. We follow individual employment histories for 1,771 initially employed and 109 initially unemployed British men from the UK Household Longitudinal Study ("Understanding Society") between 2009 and 2015. To account for unobserved heterogeneity, we apply dynamic random effect models. We find that individuals with high testosterone levels are more likely to become unemployed, but they are also more likely to exit unemployment. Based on previous studies and descriptive evidence, we argue that these effects are likely driven by personality traits and occupational sorting of men with high testosterone levels. Our findings suggest that latent biological processes can affect job search behaviour and labour market outcomes, without necessarily relating to illness and disability.
Content may be subject to copyright.
Konrad-Zuse-Strasse 1 · D-18057 Rostock · Germany · Tel +49 (0) 3 81 20 81 - 0 · Fax +49 (0) 3 81 20 81 - 202 · www.demogr.mpg.de
© Copyright is held by the authors.
Working papers of the Max Planck Institute for Demographic Research receive only limited review. Views or opinions expressed
in working papers are attributable to the authors and do not necessarily reflect those of the Institute.
MPIDR Working Paper WP 2020-033 l September 2020
https://doi.org/10.4054/MPIDR-WP-2020-033
Peter Eibich l eibich@demogr.mpg.de
Ricky Kanabar
Alexander Plum
Julian Schmied
In And Out Of Unemployment - Labour
Market Dynamics And The Role Of
Testosterone
IN AND OUT OF UNEMPLOYMENT - LABOUR MARKET DY-
NAMICS AND THE ROLE OF TESTOSTERONE
Peter Eibicha, Ricky Kanabarb, Alexander Plumc, Julian Schmieda,d
Abstract
Biological processes have provided new insights into diverging labour market trajectories. In this paper,
we use population variation in testosterone levels to explain transition probabilities into and out of un-
employment. We follow individual employment histories for 1,771 initially employed and 109 initially
 
tween 2009 and 2015. To account for unobserved heterogeneity, we apply dynamic random effect mod-
els. We find that individuals with high testosterone levels are more likely to become unemployed, but
they are also more likely to exit unemployment. Based on previous studies and descriptive evidence, we
argue that these effects are likely driven by personality traits and occupational sorting of men with high
testosterone levels. Our findings suggest that latent biological processes can affect job search behaviour
and labour market outcomes, without necessarily relating to illness and disability.
Keywords: labour market dynamics, unemployment, testosterone, random-effects probit
JEL classification: I10, J64, C23
Declarations of interest: Kanabar acknowledges generous supported from the University of Bath In-
ternational Research Funding Scheme (Developing Networks in Europe). Grant ref:VB-SP3ARK..
_____________________
a Max Planck Institute for Demographic Research, Germany
b University of Bath, United Kingdom
c New Zealand Work Research Institute, Auckland University of Technology, New Zealand
d Free University of Berlin, Germany
1 Introduction
(Arulampalam, 2001, p. 585), partly because un-
employed individuals might be perceived (and might perceive themselves) as violating a social norm.
On the other hand, it can also be a rational decision to remain unemployed for a period to hold out for a
better job offer and improve the job match. The economic literature has shown that there are various
factors which explain why individuals become unemployed or stay in unemployment. However, the
focus has been on observable factors, such as individual and household characteristics or the past un-
employment experience and duration (see, e.g., Gregg, 2001). More recent evidence points to personal-
ity traits and non-cognitive skills as influential factors of job search behaviour and unemployment du-
ration. Studies have investigated, e.g., the locus of control (Caliendo et al., 2015; Heckman et al., 2006;
Schurer, 2017), impatience (DellaVigna and Paserman, 2005), the Big 5 personality traits (Viinikainen
and Kokko, 2012), or self-efficacy and interpersonal skills (Uysal and Pohlmeier, 2011).
Hormones have been linked to a number of non-cognitive skills and personality aspects. In particular,
testosterone is prominently linked to risk-attitude and aggression (Dabbs, 1992; Dabbs et al., 2001;
Hughes and Kumari, 2019), but also to skills such as motivation, pro-social behaviour, persistence, or
numerical ability (Apicella et al., 2008; Carré and McCormick, 2008; Dabbs et al., 2001; Welker and
Carré, 2015). Likely related to these attributes, testosterone has also repeatedly been found to predict
nce (Dreher et al., 2016; Gielen et al., 2016; Nye et al., 2017). Moreover,
testosterone also seems to affect occupational choices (Dabbs, 1992; Greene et al., 2014). Yet, surpris-
ingly testosterone has not been investigated as an explanatory factor of unemployment, something we
seek to address in this paper.
We investigate whether differences in serum testosterone levels of men can explain transitions in and
out of unemployment. We use data from Understanding Society (UKHLS), a longitudinal household
survey covering about 40,000 households from the United Kingdom, and link it with the Health and
biomarkers Survey, which holds a range of biomarker data, including the circulating level of testos-
terone. We examine two samples of initially employed or initially unemployed men aged 20 to 60, and
we standardise their testosterone levels for age and time of the sample.
Taking advantage of the longitudinal nature of the data, we apply dynamic probit random-effects models
to estimate labour market transitions. Primarily, we measure the effect of testosterone levels on the
likelihood to move into a different labour force status while taking into account the prior employment
trajectory. We apply a range of random-effects models, including an extension of the Mundlak-Cham-
berlain approach to account for unobserved heterogeneity as well as the initial conditions (Wooldridge,
2005).
We contribute to the literature by providing novel evidence on latent biological mechanisms which af-
fect labour market trajectories. Previous studies have only considered inflammation markers in relation
to unemployment but not hormones such as testosterone (Sumner et al., 2020). Moreover, unlike previ-
ous studies we examine actual testosterone levels measured in a recent blood sample rather than 2D:4D
ratio, which is a prominent marker for prenatal exposure to testosterone (see, e.g., Gielen et al., 2016).
The closest study to ours is Hughes and Kumari (2019), who examined the impact of testosterone on the
likelihood of being in work for a single time point and did not take advantage of the panel structure of
the data. In contrast to our study, they did not explicitly model labour market transitions, nor did they
account for state dependence.
Findings from our preferred regression specification indicate that for unemployed men, the risk of re-
maining unemployed significantly declines in the level of testosterone. In contrast, for employed men,
the risk of becoming unemployed is higher for those with high testosterone levels. These findings are
robust against different functional specifications for testosterone and different approaches to account for
unobserved heterogeneity.
These findings might be explained in part by cognitive and non-cognitive skills that high testosterone
levels are associated with, such as numerical skills or logical reasoning. In line with previous studies,
our descriptive evidence shows that men with high testosterone levels indeed performed better in these
areas. In addition, we find suggestive evidence that individuals with higher testosterone search differ-
ently for a job.
Our findings highlight how latent biological processes (beyond illness and disease) affect labour market
outcomes. For example, when designing job search assistance programs, it is crucial for policymakers
to be aware that differences in job search behaviour can be driven by biological mechanisms. Thus, due
to their inherent personality traits, some individuals might require specific forms of assistance to thrive
(e.g., individual training rather than group sessions). This study contributes to our understanding of such
mechanisms by providing comprehensive evidence on the role of testosterone.
We proceed in Section 2 with a review of the literature on biomarkers, and we derive hypotheses based
on this literature how testosterone could translate into diverging employment histories. Section 3 pre-
sents our data, and Section 4 outlines our empirical estimation strategy. In Section 5, we show descrip-
tive evidence, and in Section 6, we present the results from our regression specifications. Section 7
discusses further possible mechanism and descriptive evidence for these potential pathways. Section 8
concludes.
2 Testosterone and the labour market
2.1 Existing literature

             


    
(2018)

(e.g., Dowd et al., 2009)
(Chandola and Zhang, 2018)

(e.g., Bann et al., 2015; Kandasamy et al., 2014)
Testosterone, in particular, has been related to different forms of health issues, e.g., cardiovascular dis-
ease (Elagizi et al., 2018), but the evidence has not been very conclusive and causal pathway not fully
understood (Bann et al., 2015; Hughes and Kumari, 2019). More convincingly, among men, testosterone
seems to affect risky health behaviours and thus, different forms of health hazards (Booth et al., 1999).
Testosterone also plays a role for demographic outcomes, such as fertility, divorce and mating (e.g.,
Bütikofer et al., 2019), fitness and sport (e.g., Hsu et al., 2015), but also for labour market outcomes
(e.g., Coates et al., 2009; Dabbs, 1992; Dabbs Jr. et al., 1990; Parslow et al., 2019).
1
For example, in a
twin study on Dutch men, more prolonged prenatal testosterone exposure led to higher earnings during
the working life (Gielen et al., 2016).
2
Other studies found education to be lower among people with
low testosterone levels (Bann et al., 2015; Nye et al., 2017). Coates and Herbert (2008) followed the
daily business of 300 traders in London and found that high levels of testosterone lead to higher profits
on that day. Testosterone also affects the choice of occupation. Low testosterone individuals seem to
choose more people-oriented jobs, whereas high testosterone individuals choose more things-oriented
jobs (Dabbs Jr. et al., 1990; Hell and Päßler, 2011; Nye and Orel, 2015).
3
Typical jobs that have been
related to high testosterone are sportsmen, sales men, actors, or politicians (Dabbs Jr. et al., 1990). The
evidence is not conclusive, though. A more robust finding is that individuals with high testosterone
levels have a higher probability to be self-employed (Greene et al., 2014; Nicolaou et al., 2017; Sapienza
et al., 2009).
1
While testosterone is present in both sexes, most of the experimental studies in the literature have focused on
men. Important exceptions looked at both sexes (Dabbs et al., 2001; Gielen et al., 2016; Nye et al., 2017; Sapienza
et al., 2009) or exclusively at women (Bütikofer et al., 2019; Parslow et al., 2019).
2
Among women, high testosterone levels are expected to be associated with higher earnings as well, as women
with higher testosterone levels tend to work in male-dominated occupations, which tend to be better paid. How-
ever, recent empirical evidence found the opposite or no effect (Bütikofer et al., 2019; Gielen et al., 2016; Nye et
al., 2017).
3
Women that have higher testosterone levels tend to choose jobs that are male-dominated, whereas women with
low levels choose more female-dominated jobs (Nye and Orel, 2015). This observation has been used to explain
parts of the gender pay gap (e.g., Gielen et al., 2016).
The findings discussed above are usually attributed to non-cognitive skills and individual characteristics
associated with high testosterone levels. Typical characteristics that have been stressed in the literature
are, among others, being independent, self-centred, adventurous, achievement-oriented, and focused on
personal goals (Greene et al., 2014). Further, high testosterone is associated with risk-taking (Apicella
et al., 2008; Coates and Herbert, 2008; Hughes and Kumari, 2019; Stenstrom et al., 2011), dominant
behaviour and aggression (Archer, 2006; Chance et al., 2000; Dabbs, 1992; Dabbs et al., 2001; Schaal
et al., 1996), but also status-enhancing pro-social behaviour.
4
For example, Dreher et al. (2016) injected
testosterone or a placebo to 40 young men and found that in an economic bargaining game, treated
individuals were indeed more aggressive towards others. Still, at the same time, they were also more
generous when it promoted social status. Similarly, individuals with high testosterone levels show more
initiative forming friendships and are, therefore, able to build up larger social networks (Booth et al.,
2006; Cheng et al., 2013). In other game studies, men with high testosterone levels were more willing
to engage in competitive tasks (Carré and McCormick, 2008) and they showed more persistence solving
an undoable task (Welker and Carré, 2015).
Cognitive abilities have also been related to testosterone. While early work reported that young boys
with high testosterone levels lack intelligence (Chance et al., 2000; Dabbs, 1992), more recent work
showed that individuals with high testosterone levels have higher numeric capabilities and thus perform
better in computer science or related occupations (Brookes et al., 2007; Brosnan et al., 2011). 
   
(Bosch-Domènech et al., 2014)

(Frederick, 2005)Finally, a series of studies showed that people with high testosterone levels perform
better in face-to-face situations (e.g., Dabbs et al., 1997; Mazur, 1985). For example, Dabbs et al. (2001)
interviewed and filmed male college students and found that individuals with high levels of testosterone
appeared more forward and independent and focused directly on the target. They were also more restless
and oriented toward action.
Given the evidence, it appears reasonable that testosterone could also explain employment dynamics.

(Arulampalam, 2001; Bijlsma et al., 2017; Marcus, 2014)

(2006)


(Schurer, 2017)(Caliendo et al., 2015)
4

(e.g., Mehta and Prasad, 2015)




2.2 Testosterone and employment transitions

 


(Archer, 2006; Chance et al., 2000)
(Dreher et al., 2016)

(Ponzi et al., 2016)


(Dabbs et al., 2001, 1997)





(Bosch-Domènech et
al., 2014)

, worried about their social status, 








    
(Brookes et al., 2007; Brosnan et
al., 2011)
   







3 Data
Understanding Society 

  Understanding Society    




     

5
  

5
The nurse health visit was conducted among adult survey participants from the General Population Sample (GPS)
which comprises of households in the UK and BHPS sample only. The nurse visit took place after wave 2 (May
2010-July 2012) for those individuals in the GPS and after wave 3 (June 2011-July 2012) for BHPS sample re-
spondents.
3.1 Health and biomarkers Survey



   (Benzeval et al.,
2014)
           


6


  Table 1   
Figure 1

7
Table 1:


























Notes   
  

6
Observational studies typically refer to levels of currently circulating testosterone, but recent studies have used
prenatal testosterone levels, approximated by physiological conditions such as the length of the second to fourth
manual digits (2D:4D) (Bütikofer et al., 2019; Coates et al., 2009; Gielen et al., 2016; Nye et al., 2017; Parslow et
al., 2019; Stenstrom et al., 2011). In contrast to current testosterone levels, prenatal exposure is expected to be
independent of early childhood environmental exposures (Nye et al., 2017). The underlying biological mechanism
is not fully understood, however.
7

Figure 1:
Notes     
  

3.2 Longitudinal data





8







  Table 2

Table 2
  


8



14 16 18 20
Mean level of testosterone (nmol/l)
20 30 40 50 60 70
age

  



Table 2:

Understanding Society



survey
-2
366
23
survey
-1
151
25
Nurse visit)
0
217
30
survey
1
369
36
survey
2
377
82
survey
3
-
-
Notes  



Table 3
5,460 observations, out of which 309 (5.7%) were unemployed during the nurse visit,
and 5,151 were employed (94.3%).
Table 3:


Employed during nurse visit


Unemployed during nurse visit



1,880
5,460
Notes

4 Methodology

(e.g., Arulampalam, 2001; Bhuller et al., 2017; Biewen and Steffes, 2010; Stewart, 2007)

genuine

(Stewart, 2007)


    
      
         

 


  
  



 
   

(1978)
(1992)
  (see also Wooldridge, 2005)

    
      
(Wooldridge, 2005)  


    

    

   

    

(see Wooldridge, 2005)

   

   
  
       
(2013)



9


  
   
    
       
  


 
 

 

   



    

  
   
    
   
  



   

 
 

 
(Butler and Moffitt, 1982)
4.1 Subsample estimations
     



  
  
    
       
        
       
  
9
Note that our findings on the effect of testosterone are robust to various specifications of including the initial
labour market status (e.g., interacting with the lagged labour market position, no interaction, not accounting for
the initial labour market status).

  
  
    
       
        
       
  
4.2 Including testosterone as a covariate


a. In the regression model, we include the absolute level of testosterone (nmol/l) as a covariate
and control for the hour of the nurse visit (Model 1). In a further specification, we include
the absolute level of testosterone as a second-degree polynomial (Model 2)
 We use the Health and biomarkers Survey to construct a sample of men with a positive level
of testosterone in the age range 16 to 70 who had their interview started between 9 am and
8 pm   . We utilise an OLS model to estimate the deviation from the time- and
age-corrected mean.
10
We order the distribution of the deviation and form three groups: (i)
low level of testosterone if the deviation belongs to the lowest decile, (ii) medium level of
testosterone if the deviation is in 2nd to 9th decile, and (iii) high level of testosterone if the
deviation belongs to the highest decile (Model 3).


4.3 Observable characteristics and individual-specific effects


(Heckman, 1981)
(Stewart, 2007)


(2005)


(2005)

given
             
10
 




(Wooldridge, 2005)


    
      
we use up to 5 observations per
individual (therefore, the number of time-points we consider is larger than in the base specification). To
account for the Mundlak specification, we specify:

  
    
  (see
also Akay, 2012; Rabe-Hesketh and Skrondal, 2013)
    
  
  
       


  
   
    
  
      



(2005)



5 Descriptive statistics


Table 4










Table 4:













1st decile




2nd – 9th decile




10th decile













Degree




Other higher degree




A-level etc




GCSE etc




Other qualication




No qualication





excellent




very good




good




fair




poor





England




Wales




Scotland















1




2




3




4+










single




married




separated, divorced, widowed




N



Notes  

Table 5
     






Table 5: 
t
t
t-1
t-1









t-1









t






Notes
  

Table 6



low testosterone is associated with
higher persistence of unemployment, while high testosterone seems to be associated with a higher risk
of entering unemployment.

Table 6:




tt-1
1st decile



2nd – 9th decile



10th decile



tt-1
1st decile



2nd – 9th decile



10th decile




  

6 Results
6.1 Base regression


Table 7

Table 7


11




 Table 7    
Table 7


11
nmol/l testosterone, there is in the beginning a
reducing effect but the slope turns positive from around 10 nmol/l.

Table 7: Effect of testosterone on unemployment risk
Full Sample
Initially unemployed
Initially employed
Model
(1)
(2)
(3)
(1)
(2)
(3)
(1)
(2)
(3)
testosterone nmol/l
0.0167
-0.0273
-0.0843*
-0.274*
0.0270*
-0.00565
(0.0127)
(0.0502)
(0.0443)
(0.165)
(0.0138)
(0.0521)
(testosterone nmol/l)2
0.00123
0.00524
0.000908
(0.00136)
(0.00417)
(0.00141)
testosterone
1st decile
reference category
2nd 9th decile
-0.132
-2.027**
0.162
(0.242)
(0.812)
(0.274)
10th decile
0.355
-2.331**
0.622*
(0.299)
(1.001)
(0.339)
Observations
5,460
5,460
5,460
309
309
309
5,151
5,151
5,151
LogLikelihood
-556
-555.6
-559.7
-93.76
-92.90
-98.25
-435.9
-435.6
-438.8
Notes  












  

Table 8: Average partial effects
Initially unemployed
Initially employed
1st decile
reference category
2nd 9th decile
-0.254*
0.004
(0.137)
(0.007)
10th decile
-0.300*
0.022
(0.168)
(0.016)
Individuals
109
1,771
Notes
***,**,* refers to statistically significant at 1%, 5% and 10% level respectively





6.2 Robustness checks


12

  



12
Additional robustness estimations which are not described in detail here include dropping covariates. However,
none of the tests lead to qualitatively different findings.



Figure A 1










Figure A 2






Table A.3 

Table 7


13



Table 7
14
6.3 Continuous (un)employment




13
Both are significant at the 1% level.
14
Significant at the 5% level.
















Table 1Table 9











Table 9: Marginal effects
Initially unemployed
Initially employed
Waves unemployed


3

1st decile
reference category
2nd 9th decile
-0.267***
-0.286**
-0.285**
0.001
(0.080)
(0.114)
(0.121)
(0.016)
10th decile
-0.169
-0.160
-0.313**
0.035
(0.140)
(0.158)
(0.145)
(0.026)
Individuals
91
91
91
1,609
Notes      ***,**,* refers to
statistically significant at 1%, 5% and 10% level respectively
7 Mechanisms
We showed that testosterone affects s transitions in and out of employment. Now we examine
whether these transitions
In section 2.2, we discussed potential channels through which testosterone may affect an individ-
uals employment status. 


For example, the numerical ability was related to testosterone in experimental studies (Brosnan et al.,
2011), and it was also collected during wave 3 in the UKHLS mainstage interview.
15
Survey respond-
ents practical numerical knowledge is assessed by testing whether they can understand percentages and
fractions in typical real-life settings. Such ability measures have been shown to be highly related to
wealth (McArdle et al., 2009; McFall, 2013). Individuals are presented with three initial problems, and
if none are answered correctly a further (simple) question is asked. On the other hand, if all questions
are answered correctly, then an additional (more difficult) question is asked; if this was also answered
correctly, a further final question is asked.
16
Thus, an individuals final score is between zero (no correct
answers) and five (all correct answers) and a clear ordering exists. Regression results (see Table A.4 in
the appendix) show that relative to individuals with a low testosterone level, the log odds of reporting a
higher test score are 1.29 times higher (significant at the 5% level) among those with a medium level of
testosterone.
17
15
This is approximately 7 months after the wave 2 nurse visit and 5 months before the wave 3 nurse visit, and
hence relatively close to the date circulating testosterone is measured.
16
This test was adopted from the English Longitudinal Study of Ageing and some parts have also been included
in the US Health and Retirement Study and Survey of Health Ageing and Retirement in Europe.
17
Low level is defined as bottom quintile of deviation from mean, medium level is between second and fourth
quintile and high level is top quintile.

Alongside numerical ability, Understanding Society assesses an individuals fluid reasoning using logic
puzzles (number series). Such measures have been found to be related to individuals financial
knowledge (Delavande et al., 2008) and are negatively associated with age (Salthouse, 2010).
18
We,
therefore, control for age in the regression analysis. Individuals in households were randomly allocated
to a set of questions.
19
Within each set, individuals were asked six questions. An individuals final
score ranged between zero (no correct answers) and six (all correct answers). Regression results Table
A.5 in the appendix) show that relative to individuals with a low testosterone level, the log odds of
reporting a higher test score are 1.28 times higher (significant at the 5% level) among those with a
medium level of testosterone. It is important to note that, given our main sample follows individuals
aged between 20 and 60 years old when initially observed, one could argue that individuals underlying
ability is relatively stable across time (as opposed to ability at very young ages). Indeed, this is one of
the underlying assumptions we make when controlling for initial conditions and time-invariant unob-
served heterogeneity following Wooldridge (2005). Thus, even though these differences in numerical
ability and fluid reasoning are in line with the literature on testosterone, they are unlikely to be the
mechanism through which testosterone affects employment status in our base model as we control for
such time-invariant unobserved heterogeneity.
20
A similar line of reasoning applies to occupational class. Given that occupational class is likely to be
time-invariant (at least in the short panel considered in this paper), it is unlikely to drive our results.
Moreover, in our data, we do not observe an association between occupational class and testosterone
levels, which earlier studies have found (Dabbs, 1992).
Research suggests that males with higher levels of testosterone are more likely to express certain per-
sonality traits and behaviours in social and professional situations (Green et al. 2014, Dabbs et al. 2001).
These same traits may help such individuals overcome adverse situations, such as unemployment. Un-
derstanding Society fielded a General Health Questionnaire at wave 3 which included attitudinal ques-
tions relating to whether individuals felt they have recently been losing confidence in themselves and,
separately, whether individuals feel they have recently been able to face up to problems. In this case,
individuals with medium levels of testosterone were significantly more likely to report a response which
suggested they were had not lost confidence or the ability to face problems, compared to individuals
with low levels of testosterone. We also consider risk-taking, which has been associated with high tes-
tosterone levels (Apicella et al., 2008; Coates and Herbert, 2008). Respondents in Understanding Society
were asked to rate their willingness to take general risks on a scale between 0 and 10, where higher
18
This test was developed for use in the US Health and Retirement Study.
19
e was a CAPI coding error
(2013) for further details.
20
Our robustness checks show that (i) our main results hold even when restricting the sample to those aged at least
the specification described in section 4.3,
and hence is time constant then our main findings hold after controlling for such unobserved factors.

values indicate a greater willingness to take risks. In a regression model controlling for age and log
earnings, we found a positive and statistically significant association between being in the high testos-
terone group and reporting a higher score (OR=1.23*).
Individuals behaviour is also strongly correlated to their personality. For example, one might expect
that individuals with higher levels of testosterone are willing to search more intensely for a job ceteris
paribus. In response to a question about job search in the last 4 weeks and asked to individuals who did
not report being in paid work in the last week or having a job, those with medium level of testosterone
were more likely to report using the internet to search for a job compared to unemployed individuals
belonging to the low testosterone group. In addition, individuals were also asked about whether they
used their network to explore employment opportunities. Based on purely descriptive evidence, the data
suggest a higher proportion of the high testosterone group mentioned such a strategy, however, this
result was not statistically significant. We also examine individuals self-reported likelihood to lose their
job in the next 12 months (very unlikely, unlikely, likely or very likely), and find a strong positive
association between those individuals who belong to the high testosterone group and the likelihood of
job loss (OR=1.37**, controlling for age and log earnings).
In summary, the associations found in Understanding Society are in line with earlier studies, showing
that testosterone is positively associated with numerical ability and cognition as well as personality traits
such as risk-taking and self-confidence. Moreover, we also find some descriptive evidence for differ-
ences in job search behaviour. While we cannot conclusively prove that these potential mechanisms
explain the observed relationship between testosterone levels and unemployment, we interpret our de-
scriptive findings as suggestive evidence that such mechanisms are likely to play a role.
8 Conclusion




  


















   
           
    













     



  


References
Akay, A., 2012. Finite-sample comparison of alternative methods for estimating dynamic panel data
models. Journal of Applied Econometrics 27, 11891204. https://doi.org/10.1002/jae.1254
Apicella, C.L., Dreber, A., Campbell, B., Gray, P.B., Hoffman, M., Little, A.C., 2008. Testosterone
and financial risk preferences. Evolution and Human Behavior 29, 384390.
https://doi.org/10.1016/j.evolhumbehav.2008.07.001
Archer, J., 2006. Testosterone and human aggression: an evaluation of the challenge hypothesis. Neu-
roscience & Biobehavioral Reviews 30, 319345. https://doi.org/10.1016/j.neubio-
rev.2004.12.007
Arulampalam, W., 2001. Is Unemployment Really Scarring? Effects of Unemployment Experiences
on Wages. The Economic Journal 111, F585F606.
Bann, D., Hardy, R., Cooper, R., Lashen, H., Keevil, B., Wu, F.C.W., Holly, J.M.P., Ong, K.K., Ben-
Shlomo, Y., Kuh, D., 2015. Socioeconomic conditions across life related to multiple measures
of the endocrine system in older adults: Longitudinal findings from a British birth cohort
study. Social Science & Medicine 147, 190199.
https://doi.org/10.1016/j.socscimed.2015.11.001
Benzeval, M., Davillas, A., Kumari, M., Lynn, P., 2014. Understanding Society: The UK Household
Longitudinal Study. Biomarker User Guide and Glossary. University of Essex, Colchester.
Bhuller, M., Brinch, C.N., Königs, S., 2017. Time Aggregation and State Dependence in Welfare Re-
ceipt. The Economic Journal 127, 18331873. https://doi.org/10.1111/ecoj.12366
Biewen, M., Steffes, S., 2010. Unemployment persistence: Is there evidence for stigma effects? Eco-
nomics Letters 106, 188190. https://doi.org/10.1016/j.econlet.2009.11.016
Bijlsma, M.J., Tarkiainen, L., Myrskylä, M., Martikainen, P., 2017. Unemployment and subsequent
depression: A mediation analysis using the parametric G-formula. Social Science & Medicine
194, 142150. https://doi.org/10.1016/j.socscimed.2017.10.011
Booth, A., Granger, D.A., Mazur, A., Kivlighan, K.T., 2006. Testosterone and Social Behavior. Social
Forces 85, 167191.

cial Behavior. Journal of Health and Social Behavior 40, 130140.
https://doi.org/10.2307/2676369
Bosch-Domènech, A., Brañas-Garza, P., Espín, A.M., 2014. Can exposure to prenatal sex hormones
(2D:4D) predict cognitive reflection? Psychoneuroendocrinology 43, 110.
https://doi.org/10.1016/j.psyneuen.2014.01.023
Brookes, H., Neave, N., Hamilton, C., Fink, B., 2007. Digit Ratio (2D:4D) and Lateralization for
Basic Numerical Quantification. Journal of Individual Differences 28, 5563.
https://doi.org/10.1027/1614-0001.28.2.55
Brosnan, M., Gallop, V., Iftikhar, N., Keogh, E., 2011. Digit ratio (2D:4D), academic performance in
computer science and computer-related anxiety. Personality and Individual Differences 51,
371375. https://doi.org/10.1016/j.paid.2010.07.009
Bütikofer, A., Figlio, D.N., Karbownik, K., Kuzawa, C.W., Salvanes, K.G., 2019. Evidence that pre-
natal testosterone transfer from male twins reduces the fertility and socioeconomic success of
their female co-twins. Proc Natl Acad Sci USA 116, 6749.
https://doi.org/10.1073/pnas.1812786116
Butler, J.S., Moffitt, R., 1982. A Computationally Efficient Quadrature Procedure for the One-Factor
Multinomial Probit Model. Econometrica 50, 761764. https://doi.org/10.2307/1912613
Caliendo, M., Cobb-Clark, D.A., Uhlendorff, A., 2015. Locus of Control and Job Search Strategies.
The Review of Economics and Statistics 97, 88103. https://doi.org/10.1162/REST_a_00459
Carré, J.M., McCormick, C.M., 2008. Aggressive behavior and change in salivary testosterone con-
centrations predict willingness to engage in a competitive task. Hormones and Behavior 54,
403409. https://doi.org/10.1016/j.yhbeh.2008.04.008
Chamberlain, G., 1992. Comment: Sequential Moment Restrictions in Panel Data. Journal of Business
& Economic Statistics 10, 2026. https://doi.org/10.2307/1391799
Chance, S.E., Brown, R.T., Dabbs, J.M., Casey, R., 2000. Testosterone, intelligence and behavior dis-
orders in young boys. Personality and Individual Differences 28, 437445.
https://doi.org/10.1016/S0191-8869(99)00110-5

Chandola, T., Zhang, N., 2018. Re-employment, job quality, health and allostatic load biomarkers:
prospective evidence from the UK Household Longitudinal Study. International Journal of Ep-
idemiology 47, 4757. https://doi.org/10.1093/ije/dyx150
Cheng, J.T., Tracy, J.L., Foulsham, T., Kingstone, A., Henrich, J., 2013. Two ways to the top: Evi-
dence that dominance and prestige are distinct yet viable avenues to social rank and influence.
Journal of Personality and Social Psychology 104, 103125. https://doi.org/10.1037/a0030398
Coates, J.M., Gurnell, M., Rustichini, A., 2009. Second-to-fourth digit ratio predicts success among
high-frequency financial traders. Proc Natl Acad Sci USA 106, 623.
https://doi.org/10.1073/pnas.0810907106
Coates, J.M., Herbert, J., 2008. Endogenous steroids and financial risk taking on a London trading
floor. Proc Natl Acad Sci USA 105, 6167. https://doi.org/10.1073/pnas.0704025105
Dabbs, J.M., 1992. Testosterone and Occupational Achievement. Social Forces 70, 813824.
https://doi.org/10.2307/2579755
Dabbs, J.M., Bernieri, F.J., Strong, R.K., Campo, R., Milun, R., 2001. Going on Stage: Testosterone in
Greetings and Meetings. Journal of Research in Personality 35, 2740.
https://doi.org/10.1006/jrpe.2001.2305
Dabbs, J.M., Strong, R., Milun, R., 1997. Exploring the Mind of Testosterone: A Beeper Study. Jour-
nal of Research in Personality 31, 577587. https://doi.org/10.1006/jrpe.1997.2203
Dabbs Jr., J.M., de la Rue, D., Williams, P.M., 1990. Testosterone and occupational choice: Actors,
ministers, and other men. Journal of Personality and Social Psychology 59, 12611265.
https://doi.org/10.1037/0022-3514.59.6.1261
Delavande, A., Rohwedder, S., Willis, R., 2008. Preparation for Retirement, Financial Literacy and
Cognitive Resources. University of Michigan, Michigan Retirement Research Center.
DellaVigna, S., Paserman, M.D., 2005. Job Search and Impatience. Journal of Labor Economics 23,
527588. https://doi.org/10.1086/430286
Dowd, J.B., Zajacova, A., Aiello, A., 2009. Early origins of health disparities: Burden of infection,
health, and socioeconomic status in U.S. children. Social Science & Medicine 68, 699707.
https://doi.org/10.1016/j.socscimed.2008.12.010
Dreher, J.-2016. Testosterone
causes both prosocial and antisocial status-enhancing behaviors in human males. Proc Natl
Acad Sci USA 113, 11633. https://doi.org/10.1073/pnas.1608085113
Elagizi, A., Köhler, T.S., Lavie, C.J., 2018. Testosterone and Cardiovascular Health. Mayo Clinic Pro-
ceedings 93, 83100. https://doi.org/10.1016/j.mayocp.2017.11.006
Frederick, S., 2005. Cognitive Reflection and Decision Making. Journal of Economic Perspectives 19,
2542. https://doi.org/10.1257/089533005775196732
Gielen, A.C., Holmes, J., Myers, C., 2016. Prenatal Testosterone and the Earnings of Men and
Women. Journal of Human Resources 51, 3061. https://doi.org/10.3368/jhr.51.1.30
Greene, F.J., Han, L., Martin, S., Zhang, S., Wittert, G., 2014. Testosterone is associated with self-em-
ployment among Australian men. Economics & Human Biology 13, 7684.
https://doi.org/10.1016/j.ehb.2013.02.003
Gregg, P., 2001. The Impact of Youth Unemployment on Adult Unemployment in the NCDS. The
Economic Journal 111, 626653. https://doi.org/10.1111/1468-0297.00666
Harris, K.M., Schorpp, K.M., 2018. Integrating Biomarkers in Social Stratification and Health Re-
search. Annu Rev Sociol 44, 361386. https://doi.org/10.1146/annurev-soc-060116-053339
Heckman, J.J., 1981. The Incidental Parameters Problem and the Problem of Initial Conditions in Esti-
mating a Discrete Time-Discrete Data Stochastic Processs, in: Manski, C.F., McFadden, D.
(Eds.), Structural Analysis of Discrete Data with Econometric Applications. MIT Press, Cam-
bridge, Massachusetts, pp. 179195.
Heckman, J.J., Stixrud, J., Urzua, S., 2006. The Effects of Cognitive and Noncognitive Abilities on
Labor Market Outcomes and Social Behavior. Journal of Labor Economics 24,
411482. https://doi.org/10.1086/504455
Hell, B., Päßler, K., 2011. Are occupational interests hormonally influenced? The 2D:4D-interest
nexus. Personality and Individual Differences 51, 376380.
https://doi.org/10.1016/j.paid.2010.05.033

Hsu, C.-C., Su, B., Kan, N.-W., Lai, S.-L., Fong, T.-H., Chi, C.-P., Chang, C.-C., Hsu, M.-C., 2015.
Elite Collegiate Tennis Athletes Have Lower 2D: 4D Ratios Than Those of Nonathlete Con-
trols. The Journal of Strength & Conditioning Research 29.
Hughes, A., Kumari, M., 2019. Testosterone, risk, and socioeconomic position in British men: Explor-
ing causal directionality. Social Science & Medicine 220, 129140.
https://doi.org/10.1016/j.socscimed.2018.11.004
Kandasamy, N., Hardy, B., Page, L., Schaffner, M., Graggaber, J., Powlson, A.S., Fletcher, P.C., Gur-
nell, M., Coates, J., 2014. Cortisol shifts financial risk preferences. Proc Natl Acad Sci USA
111, 3608. https://doi.org/10.1073/pnas.1317908111
Kroft, K., Lange, F., Notowidigdo, M.J., 2013. Duration Dependence and Labor Market Conditions:
Evidence from a Field Experiment*. The Quarterly Journal of Economics 128, 11231167.
https://doi.org/10.1093/qje/qjt015
Marcus, J., 2014. Does Job Loss Make You Smoke and Gain Weight? Economica 81, 626648.
https://doi.org/10.1111/ecca.12095
Mazur, A., 1985. A biosocial model of status in face-to-face primate groups. Social Forces 64, 377
402. https://doi.org/10.2307/2578647
McArdle, J.J., Smith, J.P., Willis, R., 2009. Cognition and Economic Outcomes in the Health and Re-
tirement Survey. National Bureau of Economic Research Working Paper Series No. 15266.
https://doi.org/10.3386/w15266
McFall, S., 2013. Understanding Society- UK Household Longitudinal Study: Cognitive Ability
Measures (Understanding Society User Manual).
Mehta, P.H., Prasad, S., 2015. The dual-hormone hypothesis: a brief review and future research
agenda. Current Opinion in Behavioral Sciences 3, 163168.
https://doi.org/10.1016/j.cobeha.2015.04.008
Mundlak, Y., 1978. On the Pooling of Time Series and Cross Section Data. Econometrica 46, 6985.
https://doi.org/10.2307/1913646
Nicolaou, N., Patel, P.C., Wolfe, M.T., 2017. Testosterone and Tendency to Engage in Self-Employ-
ment. Management Science 64, 18251841. https://doi.org/10.1287/mnsc.2016.2664
Nye, J.V.C., Bryukhanov, M., Kochergina, E., Orel, E., Polyachenko, S., Yudkevich, M., 2017. The
effects of prenatal testosterone on wages: Evidence from Russia. Economics & Human Biol-
ogy 24, 4360. https://doi.org/10.1016/j.ehb.2016.11.003
Nye, J.V.C., Orel, E., 2015. The influence of prenatal hormones on occupational choice: 2D:4D evi-
dence from Moscow. Personality and Individual Differences 78, 3942.
https://doi.org/10.1016/j.paid.2015.01.016
Parslow, E., Ranehill, E., Zethraeus, N., Blomberg, L., von Schoultz, B., Hirschberg, A.L., Johannes-
son, M., Dreber, A., 2019. The digit ratio (2D:4D) and economic preferences: no robust asso-
ciations in a sample of 330 women. Journal of the Economic Science Association 5, 149169.
https://doi.org/10.1007/s40881-019-00076-y
Ponzi, D., Zilioli, S., Mehta, P.H., Maslov, A., Watson, N.V., 2016. Social network centrality and hor-
mones: The interaction of testosterone and cortisol. Psychoneuroendocrinology 68, 613.
https://doi.org/10.1016/j.psyneuen.2016.02.014
Rabe-mple solution to
the initial conditions problem. Economics Letters 120, 346349.
https://doi.org/10.1016/j.econlet.2013.05.009
Salthouse, T.A., 2010. Selective review of cognitive aging. Journal of the International Neuropsycho-
logical Society 16, 754760. https://doi.org/10.1017/S1355617710000706
Sapienza, P., Zingales, L., Maestripieri, D., 2009. Gender differences in financial risk aversion and ca-
reer choices are affected by testosterone. Proc Natl Acad Sci USA 106, 15268.
https://doi.org/10.1073/pnas.0907352106
Schaal, B., Tremblay, R.E., Soussignan, R., Susman, E.J., 1996. Male Testosterone Linked to High
Social Dominance but Low Physical Aggression in Early Adolescence. Journal of the Ameri-
can Academy of Child & Adolescent Psychiatry 35, 13221330.
https://doi.org/10.1097/00004583-199610000-00019
Schurer, S., 2017. Bouncing back from health shocks: Locus of control and labor supply. Journal of
Economic Behavior & Organization 133, 120. https://doi.org/10.1016/j.jebo.2016.10.019

Stenstrom, E., Saad, G., Nepomuceno, M.V., Mendenhall, Z., 2011. Testosterone and domain-specific
risk: Digit ratios (2D:4D and rel2) as predictors of recreational, financial, and social risk-tak-
ing behaviors. Personality and Individual Differences 51, 412416.
https://doi.org/10.1016/j.paid.2010.07.003
Stewart, M.B., 2007. The interrelated dynamics of unemployment and low-wage employment. Journal
of Applied Econometrics 22, 511531. https://doi.org/10.1002/jae.922
Sumner, R.C., Bennett, R., Creaven, A.-M., Gallagher, S., 2020. Unemployment, employment precar-
ity, and inflammation. Brain, Behavior, and Immunity 83, 303308.
https://doi.org/10.1016/j.bbi.2019.10.013
Uysal, S.D., Pohlmeier, W., 2011. Unemployment duration and personality. Journal of Economic Psy-
chology 32, 980992. https://doi.org/10.1016/j.joep.2011.03.008
Viinikainen, J., Kokko, K., 2012. Personality traits and unemployment: Evidence from longitudinal
data. Journal of Economic Psychology 33, 12041222.
https://doi.org/10.1016/j.joep.2012.09.001
Welker, K.M., Carré, J.M., 2015. Individual Differences in Testosterone Predict Persistence in Men.
European Journal of Personality 29, 8389. https://doi.org/10.1002/per.1958
Wooldridge, J.M., 2005. Simple solutions to the initial conditions problem in dynamic, nonlinear
panel data models with unobserved heterogeneity. Journal of Applied Econometrics 20, 39
54. https://doi.org/10.1002/jae.770


Figure A 1: Robustness check for different age ranges
-6 -4 -2 0
20-60
21-59
22-58
23-57
24-56
25-55
20-60
21-59
22-58
23-57
24-56
25-55
2nd – 9th decile 10th decile
testosterone
age window
Initially unemployed
-.5 0.5 11.5
20-60
21-59
22-58
23-57
24-56
25-55
20-60
21-59
22-58
23-57
24-56
25-55
2nd – 9th decile 10th decile
testosterone
age window
Initially employed

Figure A 2: Robustness check for different testosterone cut-off points
-6 -4 -2 0
5 - 95
6 - 94
7 - 93
8 - 92
9 - 91
10 - 90
11 - 89
12 - 88
13 - 87
14 - 86
15 - 85
16 - 84
17 - 83
18 - 82
19 - 81
20 - 80
5 - 95
6 - 94
7 - 93
8 - 92
9 - 91
10 - 90
11 - 89
12 - 88
13 - 87
14 - 86
15 - 85
16 - 84
17 - 83
18 - 82
19 - 81
20 - 80
Medium range Top range
testosterone
cut-off points
Initially unemployed
-1 012
5 - 95
6 - 94
7 - 93
8 - 92
9 - 91
10 - 90
11 - 89
12 - 88
13 - 87
14 - 86
15 - 85
16 - 84
17 - 83
18 - 82
19 - 81
20 - 80
5 - 95
6 - 94
7 - 93
8 - 92
9 - 91
10 - 90
11 - 89
12 - 88
13 - 87
14 - 86
15 - 85
16 - 84
17 - 83
18 - 82
19 - 81
20 - 80
Medium range Top range
testosterone
cut-off points
Initially employed


Initially unemployed
Initially employed
Specification
testosterone
1st decile
2nd 9th decile
-4.13***
0.4
(1.47)
(0.43)
10th decile
-3.21**
1.18**
(1.42)
(0.58)
Observations
371
6,408
LogLikelihood
-50.59
-297.4
Notes: own calculations using data from Understanding Society
subsample Health and biomarkers Survey. N= 6,771.
***,**,* refers to statistically significant at 1%, 5% and 10% level respectively

Total score: numerical ability
Specification
Testosterone group
Low (1st quintile)
Reference group
Medium (2nd- 4th quintile)
1.29***
[0.10]
High (5th quintile)
1.01
[0.003]
Age
1.01***
[0.003]
Observations
3,123
LogLikelihood
-3844.41
Notes: own calculations using data from Understanding Society
***,**,* refers to statistically significant at 1%, 5% and 10% level respec-



Total score: fluid reasoning
Specification
Testosterone group
Low (1st quintile)
Reference group
Medium (2nd- 4th quintile)
1.28**
[0.14]
High (5th quintile)
1.22
[0.18]
Age
0.99***
[0.004]
Observations
1,540
LogLikelihood
-2358.37
Notes: own calculations using data from Understanding Society
***,**,* refers to statistically significant at 1%, 5% and 10% level respec-
tively. Coefficients refer to odds ratio .
... Positive associations with circulating testosterone have for example been reported for selfemployment, a financially 'riskier' strategy than standard employment (9,10), and likelihood of employment transitions (11). Work in male occupational samples points to a positive relationship of testosterone with aspects of socioeconomic position: among male executives, circulating testosterone has been linked with number of subordinates (12), and among male financial traders, with daily profits (13). ...
Full-text available
Preprint
Men with more advantaged socioeconomic position (SEP) and better health have been observed to have higher levels of testosterone. It is unclear whether these associations arise because testosterone has a causal impact on SEP and health. In 306,248 participants of UK Biobank, we performed sex- stratified genome-wide association analysis to identify genetic variants associated with testosterone. Using the identified variants, we performed Mendelian randomization analysis of the influence of testosterone on socioeconomic position, including income, employment status, area-level deprivation, and educational qualifications; on health, including self-rated health and BMI, and on risk-taking behaviour. We found little evidence that testosterone affected socioeconomic position, health, or risk-taking. Our results therefore suggest it is unlikely that testosterone meaningfully affects these outcomes in men or women. Differences between Mendelian randomization and multivariable-adjusted estimates suggest previously reported associations with socioeconomic position and health may be due to residual confounding or reverse causation.
Full-text available
Article
Unemployment has been associated with poorer health, but few studies have examined the biological mechanisms that confer these health decrements. Further, no studies to date have examined differences across employment groups to consider whether employment (in whatever means) is preferential in terms of health. The present study utilised secondary data from Understanding Society: The Household Longitudinal Survey during the aftermath of the recent global recession. Two markers of peripheral inflammation: C-reactive protein (CRP) and fibrinogen were assessed across employment groups (unemployed; permanent, temporary, and self-employed), controlling for individual, socio-demographic and health variables to give greater context to our understanding of how employment status influences health. After controlling for relevant confounds, unemployment was associated with higher levels of fibrinogen but not CRP. Subsequent analyses of employment subgroup revealed the temporary employed have similar levels of fibrinogen to the unemployed, and may therefore be at a similar health disadvantage. The findings confirm that unemployment is associated with increases in one marker of peripheral inflammation, but that this health protection is not conferred to those in precarious employment.
Full-text available
Article
Many studies report on the association between 2D:4D, a putative marker for prenatal testosterone exposure, and economic preferences. However, most of these studies have limited sample sizes and test multiple hypotheses (without preregistration). In this study we mainly replicate the common specifications found in the literature for the association between the 2D:4D ratio and risk taking, the willingness to compete, and dictator game giving separately. In a sample of 330 women we find no robust associations between any of these economic preferences and 2D:4D. We find no evidence of a statistically significant relation for 16 of the 18 total regressions we run. The two regression specifications which are statistically significant have not previously been reported and the associations are not in the expected direction, and therefore they are unlikely to represent a real effect.
Full-text available
Article
Significance In litter-bearing species, females exposed to prenatal testosterone from male littermates exhibit altered traits. In humans, rising twinning rates may be exposing a growing subset of females to similar effects. Data on all twin births in Norway between 1967 and 1978 show that females exposed in utero to a male co-twin have a decreased probability of graduating from high school (15.2%), completing college (3.9%), and being married (11.7%), and have lower fertility (5.8%) and life-cycle earnings (8.6%). These relationships remain unchanged among females whose male co-twin died soon after birth, implicating prenatal testosterone exposure rather than being raised with a male sibling. These findings support the hypothesis that being exposed to a male co-twin in utero can have lasting effects on females.
Full-text available
Article
Lower testosterone levels in men are observationally associated with worse health, but it is unclear whether they contribute to well-established social gradients in health. Mendelian Randomization studies suggest positive testosterone-health associations may not be causal, with some intervention studies suggesting testosterone administration could be harmful. Since testosterone is rarely measured in general population studies, very little is known about how testosterone varies by social position. Differences by education and household income in British men aged 60-64y were recently reported, but it is unclear whether this reflects an influence of socioeconomic position (SEP) on testosterone, influence of testosterone on SEP, or confounding. In the UK Household Longitudinal Study, a nationally-representative survey of UK adults, we examine social differences in testosterone in 3663 men aged 16-97y in 2010–12. We consider diverse dimensions of SEP: education, employment status, equivalized household income and personal earnings. Multivariable regression is used to explore social differences in testosterone across the adult life-span (16-97y). Secondly, Mendelian Randomization (MR), an approach which uses gene variants as instrumental variables for endogenous exposures, is used to investigate causal directionality. We examine associations with risk-taking, a plausible mediator of testosterone-SEP associations. In observational models no social differences in testosterone are seen, but MR models suggest a positive influence of testosterone on earnings (increase in log-transformed monthly earnings (GBP) per standard deviation increase in testosterone: 0.51, 95%CI: 0.03,1.05, p = 0.07) and probability of being in work (probit coefficient:0.25, 95%CI: 0.01,0.51, p = 0.06). Though MR estimates are less precise, results are consistent with previous literature linking testosterone with labour market success. The discrepancy may reflect suppression of observational associations by factors positively correlated with testosterone and negatively correlated with SEP, or indicate an influence of typical lifetime testosterone, which may be better indexed by genetic variants than by single testosterone measurements subject to noise.
Full-text available
Article
Background: There is little evidence on whether becoming re-employed in poor quality work is better for health and well-being than remaining unemployed. We examined associations of job transition with health and chronic stress-related biomarkers among a population-representative cohort of unemployed British adults. Methods: A prospective cohort of 1116 eligible participants aged 35 to 75 years, who were unemployed at wave 1 (2009/10) of the UK Household Longitudinal Study, were followed up at waves 2 (2010/11) and 3 (2011/12) for allostatic load biomarkers and self-reported health. Negative binomial and multiple regression models estimated the association between job adversity and these outcomes. Results: Compared with adults who remained unemployed, formerly unemployed adults who transitioned into poor quality jobs had higher levels of overall allostatic load (0.51, 0.32-0.71), log HbA1c (0.06, <0.001-0.12), log triglycerides (0.39, 0.22-0.56), log C-reactive protein (0.45, 0.16-0.75), log fibrinogen (0.09, 0.01-0.17) and total cholesterol to high-density lipoprotein (HDL) ratio (1.38, 0.88-1.88). Moreover, physically healthier respondents at wave 1 were more likely to transition into good quality and poor quality jobs after 1 year than those who remained unemployed. Conclusions: Formerly unemployed adults who transitioned into poor quality work had greater adverse levels of biomarkers compared with their peers who remained unemployed. The selection of healthier unemployed adults into these poor quality or stressful jobs was unlikely to explain their elevated levels of chronic stress-related biomarkers. Job quality cannot be disregarded from the employment success of the unemployed, and may have important implications for their health and well-being.
Article
This article provides an overview of the integration of biomarkers and biological mechanisms in social science models of stratification and health. The goal in reviewing this literature is to highlight research that identifies the social forces that drive inequalities over the life course and across generations. The article is structured in the following way. First, descriptive background information on biomarkers is presented, and second, the general theoretical paradigms that lend themselves to an integrative approach are reviewed. Third, the biomarkers used to capture several biological systems that are most responsive to social conditions are described. Fourth, research that explicates how social exposures "get under the skin" to affect physiological functioning and downstream health is discussed, using socioeconomic disadvantage as an illustrative social exposure. The review ends with emerging directions in the use of biomarkers in social science research. This article endeavors to encourage sociologists to embrace biosocial approaches in order to elevate the importance of social factors in biomedical processes and to intervene on the social conditions that create unjust and avoidable inequities.
Article
There is an ongoing debate in the medical community regarding the effects of testosterone on cardiovascular (CV) health. For decades, there has been conflicting evidence regarding the association of endogenous testosterone levels and CV disease (CVD) events that has resulted in much debate and confusion among health care providers and patients alike. Testosterone therapy has become increasingly widespread, and after the emergence of studies that reported increased CVD events in patients receiving testosterone therapy, the US Food and Drug Administration (FDA) released a warning statement about testosterone and its potential risk regarding CV health. Some of these studies were later found to be critically flawed, and some experts, including the American Association of Clinical Endocrinologists and an expert panel regarding testosterone deficiency and its treatment, reported that some of the FDA statements regarding testosterone therapy were lacking scientific evidence. This article summarizes the current evidence regarding the relationship between testosterone (endogenous and supplemental) and CV health. A literature review was conducted via search using PubMed and specific journal databases, including the New England Journal of Medicine and the Journal of the American College of Cardiology. Key search terms included testosterone and cardiovascular health, coronary artery disease, heart failure, androgen deprivation therapy, intima-media thickness, and adrenal androgens. Initial study selection was limited to publications within the past 10 years (January 1, 2007, through December 31, 2016); however, key publications outside of this time frame were selected if they provided important quantitative data or historical perspectives for the review of this topic. The search was further supplemented by reviewing references in selected articles.
Article
The effects of unemployment on depression are difficult to establish because of confounding and limited understanding of the mechanisms at the population level. In particular, due to longitudinal interdependencies between exposures, mediators and outcomes, intermediate confounding is an obstacle for mediation analyses. Using longitudinal Finnish register data on socio-economic characteristics and medication purchases, we extracted individuals who entered the labor market between ages 16 and 25 in the period 1996 to 2001 and followed them until the year 2007 (n = 42,172). With the parametric G-formula we estimated the population-averaged effect on first antidepressant purchase of a simulated intervention which set all unemployed person-years to employed. In the data, 74% of person-years were employed and 8% unemployed, the rest belonging to studying or other status. In the intervention scenario, employment rose to 85% and the hazard of first antidepressant purchase decreased by 7.6%. Of this reduction 61% was mediated, operating primarily through changes in income and household status, while mediation through other health conditions was negligible. These effects were negligible for women and particularly prominent among less educated men. By taking complex interdependencies into account in a framework of observed repeated measures data, we found that eradicating unemployment raises income levels, promotes family formation, and thereby reduces antidepressant consumption at the population-level.
Article
Does testosterone increase the tendency to engage in self-employment? The results presented to date have been mixed. Using three different studies, we provide additional evidence on the relationship between testosterone and self-employment. Drawing on a cross section of 2,146 individuals (1,178 males and 968 females) from the National Health and Nutrition Examination Surveys’ 2011–2012 sample, and controlling for endogeneity (with red blood cell count, percentage hematocrit, and zinc supplement intake in the past 30 days as instruments), we find that serum testosterone levels are positively associated with self-employment for males (marginally significant, two-tailed test). As testosterone levels could be affected by social, economic, and biological factors during one’s life course, we draw more robust inferences by assessing whether the 2D:4D digit ratio, a marker of prenatal testosterone exposure, influences the likelihood of self-employment. From Understanding Society’s Innovation Panel Wave 6, we tested separate models for 449 males and 525 females, and our results indicate that males (respectively, females) with a lower 2D:4D ratio in their left hand, or higher prenatal testosterone exposure, have a significantly greater (respectively, marginally significant) likelihood of self-employment (two-tailed test). Finally, we examine the twin testosterone transfer effect in a sample of opposite-sex and same-sex twins from the National Survey of Midlife Development in the United States and provide additional support for the marginally significant (two-tailed test) positive association between testosterone and self-employment. Data are available at https://doi.org/10.1287/mnsc.2016.2664 . This paper was accepted by Toby Stuart, entrepreneurship and innovation.
Article
Is in utero exposure to testosterone correlated with earnings? The question matters for understanding determinants of wage differences that have attracted so much attention among economists in the past decade. Evidence indicates that markers for early testosterone exposure are correlated with traits like risk-taking and aggressiveness. But it is not at all clear how such findings might map into labor market success. We combine unique data from the Russian Longitudinal Monitoring Survey with measured markers (2D:4D ratios) for testosterone exposure and find that lower digit ratios (higher T) correlate with higher wages for women and for men, when controlling for age, education and occupation. There is also some evidence of a potential non-linear, inverse U-effect of digit ratios on wages but this is sensitive to choice of specification. These findings are consistent with earlier work on prenatal T and success in careers (Coates et al., 2009) but inconsistent with the work of Gielen et al. (2016) who find differing effects for men and women.