JUNE 2022

Hiring Bias
Employment for Muslim women at entry-level roles

Authors:
Dr. Ruha Shadab
Vanshika Sharan
Deepanjali Lahiri
TABLE OF
CONTENTS
Executive Summary 01

Introduction 02

Literature Review 03

Methodology 05

Limitations 06

Data and Findings 07

Conclusion 11

Solutions and Recommendations 12

Conflict of Interest 13

Acknowledgment 14

References 15
EXECUTIVE
SUMMARY
Muslim women receive only half the job call-backs that Hindu
women receive

India has witnessed a sharp decline in women’s participation in the workforce over the past two decades.
Limited female participation has different root causes depending on the intersectionality of identities. This
research attempts to understand one of the challenges Muslim women face as they enter the workforce by
studying the differences in recruiter response rates between Hindu and Muslim women applicants.

Methodology

The research took an experimental approach and collected primary data through correspondence or resume
study. Two equally qualified profiles were created to match the market standard for entry-level roles in India.
The only variable was the names; Habiba Ali for the Muslim profile and Priyanka Sharma for the Hindu. The
profiles were created without incorporating photographs. Over ten months, 2,000 job applications were sent
through 1,000 job postings on job search sites like LinkedIn and Naukri.com. The responses were collected
and used to ascertain a net discrimination rate.

Findings

1. The net discrimination rate was 47.1%, as the Hindu woman received 208 positive responses, while the
Muslim woman received half of that (103). This was evident across industries.
2. Recruiters were more cordial to the Hindu candidate; 41.3% of the recruiters had connected with Priyanka
over phone calls, while only 12.6% spoke with Habiba over a call.
3. North India had a lower discrimination rate (40%) compared to jobs located in West (59%) and South India
(60%).

Recommendations

India’s economy is missing out on talent. Organizations, researchers, and individuals all can play a role in
creating an ecosystem that supports Muslim women with employment.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 1

INTRODUCTION
Although financial independence is crucial, it remains asymmetric
in its distribution to all women. Women who lie at an intersection of
subaltern identities suffer dual oppression and cannot access the
same opportunities. This asymmetry is felt heavily by Muslim
women, who are incredibly underrepresented in the workforce.

Women empowerment and economic standing are inextricably linked (Duflo, 2012). Studies show that due to
traditional gender roles, women find themselves disproportionately engaged in work that is confined to the
home. Women spend twice as much time on housework, five times as much on child care, and half as much
time on market work as men (Berniell and Sánchez-Páramo, 2011) facilitating a system where women are
largely financially dependent on the men of the house. The total number of women employed in the labour
force in India is 149.8 million (Census, 2011), which translates to only a quarter of the country’s workforce, a
percentage that has been on a steady decline. Access to employment dictates the ability of women to
disentangle themselves from the norms that the patriarchal home assigns to them.

Although financial independence is crucial, it remains asymmetric in its distribution to all women. Women who
lie at an intersection of subaltern identities suffer dual oppression and cannot access the same opportunities.
This asymmetry is felt heavily by Muslim women, who are incredibly underrepresented in the workforce.
According to the 66th round of the National Sample Survey Organisation (2009-10), out of every 1000 working
women, only 101 were Muslim, with their share of the workforce being a meager 10%. As per the 2011 Census,
50 million Muslim women are of working age (15-65). According to a census study carried out by the
government and quoted by Mukhtar Abbas Naqvi in 2016, the Indian worker-population ratio of Muslims was
found to be the lowest relative to other communities at 32.6. The report also pointed out that the low work
participation of Muslim women is a significant reason why Muslims lag in employment figures concerning their
population, with the proportion of Muslim women working within the home being at 70% compared to the
national average of 51% for other communities.

The government has made illegal the more extreme forms of discrimination, which leads to many believing
that disparities and inequality are a facet of the past (Pager, 2007). Within the job market, hiring discrimination
is illegal and often against the belief systems that companies espouse. Consequently, many believe that
when there are disparities between communities in the workforce, it is instead the consequence of a lack of
academic ability or experience rather than discrimination within the hiring process. This more subtle but
pernicious form of discrimination becomes challenging to measure and prove. This study aims to look at the
comparison between two identically qualified candidates, with the only variable in question being their
religious background, to see if opportunities are symmetrically accessible to Hindu women and Muslim
women.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 2

LITERATURE
REVIEW
Research around the broad domain of ‘inequality’ has been far from scarce. Economic, sociological, and
anthropological research often seeks to look at the dynamics of subaltern groups within a society, with their
financial position being a crucial point of study. To fully situate this research within the existing literature, it is
vital to understand the gaps in research pertaining to Muslim women and their socioeconomic status, and the
intricacies of previous resume/correspondence studies.

India is a fragmented country with a high degree of social stratification that corresponds to a hierarchy of
power and resource (Gupta, 1992). Constitutional frameworks applied to inequality look at policy changes and
the question that persists is why inequality exists post-legal equality. Collins (1979) talks of the enduring
impacts of these hierarchies on economic opportunities in the US. Certain high-paying jobs get monopolized
by the communities with the most amount of privilege and access.

Lawmakers have made an attempt to bridge disparities by outlawing exclusion, giving rights, and attempting
to redress grievances (Darity & Deshpande, 2000). This allows people to assume discrimination is relegated
to the past, which is helped by the difficulty in detecting and measuring the subtler impacts of discrimination
(Pager, 2007). Hiring biases are thus a larger institutional problem than the result of individual biases.
Sociologically, cognitive biases hold an elevated role in the exclusion of subaltern groups from participation
in the labour market (Massey, 2007).

Much of the research around market disparities focuses on descriptive statistics from surveys, government
data around atrocities, qualitative fieldwork, and media accounts. Deshpande’s research using the National
Sample Surveys (2003) shows us that there has been no real decline in employment discrimination except for
in the case of OBC individuals. The type of data available specifically for Indian Muslims is the same as it is for
all subaltern communities, but it is insufficient in studying hiring biases as quantitative evidence is not
multivariate. Prior to Thorat and Attewell’s (2007) seminal study on hiring biases, there were little to no
studies that separated human capital differences from discrimination in the hiring process.

Resume studies or correspondence studies are ones where two or more identical profiles are developed with
one or a few variables being studied. These profiles are used to apply to jobs and by looking at the difference
in the number of responses, labour market discrimination is confirmed and quantified. Since 2009, with the
advent of job search websites, the number of correspondence studies has increased (Adamovic, 2020).
Resume studies have emerged as a more detailed and accurate measure to analyze labour market
discrimination (Bertrand & Duflo, 2016; Carlsson & Rooth, 2015; Rooth, 2014). These studies are uniquely
beneficial as it measures the discrimination during the initial phases of the recruitment process (Riach & Rich,
2002).

As per Adamovic (2020), the total count of investigated variables was 286. Several resume studies
investigated interactive effects (e.g., ethnicity and gender: Andriessen, Nievers, Dagevos, & Faulk, 2012; Edo
et al., 2017; Midtboen, 2016). Prior resume studies have analysed a large variety of variables. Most
researchers have focused on ethnic discrimination (123), followed by gender (32) and age discrimination (22).
Some studies have analysed additional diversity attributes like sexual orientation, disability, religion, socio-
economic status, and physical appearance.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 3

In India, Thorat and Attewell (2007) focused on entry-level or near entry-level jobs and sought to look at the
hiring biases between Muslim, Dalit, and upper-caste Hindu men. They applied for 548 jobs over 66 weeks
and found that the odds of a positive outcome for the Muslim name was 0.33 and for the Dalit name was 0.67
of an equivalently qualified upper-caste Hindu name. While this research proves that there is discrimination
between upper-caste Hindu men and Muslim men, it does not venture into whether those same biases exist
between upper-caste Hindu women and Muslim women. This gap still remains in the literature around hiring
discrimination today.

Muslim women are one of the most backward educationally, socially, economically, and health-wise
according to many studies (Kazi, 1999; Saaed, 2005; Shukla, 1996). The literacy rate of Muslim women was
found to be 51.9%, lower than even the poor national average for women of 65.46% (Census, 2011). Their
participation in public life and the labour market is disproportionately low. Most of the literature on the
marginalisation of Muslim women focuses on personal law and constitutional frameworks (Shukla 1996;
Ahmed and Singh 2012; Subramanian 2008; Suneetha 2012) rather than on their presence in the labour force.
While these disparities are very telling, there remains no conclusive answer on whether hiring biases are part
of why Muslim women are not proportionately represented in the labour market, which is the question that
this study seeks to answer.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 4

METHODOLOGY
This research took an experimental approach and collected primary data through correspondence or resume
study. An advantage of conducting a resume study is that it measures discrimination in a real work
environment (Bertrand & Mullainathan, 2004), which represents a crucial advantage over laboratory studies
that can only analyse artificial settings. It also allows the researchers to keep all other variables constant but
the variable of interest, making the research very versatile.

Our goal was to ascertain a net discrimination rate to compare the responses given to Hindu and Muslim
women in India. The net discrimination rate provides a percentage to illustrate the cases where only the
majority of group member has been invited for a job interview. Net discrimination rate circumvents the
problem with unobserved individual heterogeneity – a common problem in studies using administrative data
(Carlsson et al., 2014). Admittedly, an issue with resume studies is that they cannot discern between
preference-based (Becker, 1957) and statistical discrimination (Aigner & Cain, 1977; Arrow, 1973; Phelps,
1972), where preference-based discrimination refers to discrimination done on the basis of prejudice where a
recruiter’s individual biases lead to discrimination and statistical discrimination done on the basis of risk
aversion where the recruiter refrains from hiring an individual to minimize risk costs. That being said, the
inability to separate between these two is not a huge drawback unless the aim is solely to identify one of the
two (Carlsson et al., 2014), which is not the purpose of this research.

The research focused only on entry-level jobs in India in the vein of Thorat and Attewell’s study (2007) that
have a large volume of applicants and job postings. This served two purposes - it allowed the study to remain
undetected, which is crucial to getting accurate results, and allowed us to identify a large pool of similar jobs.
Applications were made to roles that have significant female participation - HR, Social Media, Digital
Marketing, etc. As has been the case with resume studies post-2009 (Adamovic, 2020), the data collection
was done through job search websites like Naukri.com along with direct applications and applications
through LinkedIn to allow us to cast a wider net. Resume studies are an effective and accurate way to
measure hiring discrimination, but it was vital to attempt to reduce error to the highest degree possible, a
crucial consideration taken into account when devising a methodology for this study. Measures were taken
including making realistic, equally-qualified dummy profiles and rotating the order of application. These
profiles did not include photos so as to keep other factors such as skin colour, clothing, and so on from
impacting the hiring process. Halfway through the study, the two resumes were also switched to minimise the
possibility of other factors causing the disparity.

Older studies used sample sizes around 500 (Attewell & Thorat, 2007), but over the past decade, this size
has ballooned to 1000 and sometimes even more. Adamovic (2020), in his review of resume studies and
formulation of an ideal methodology, states that researchers need to send out more job applications if they
have a higher number of job candidates. Considering the fact that this report is studying a singular variable,
the study utilised a standard sample size of 1000.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 5

LIMITATIONS OF
THE STUDY
There are two broad challenges - detection by the companies being studied and other variables interfering
with the results. The first issue has been addressed in two ways. As previously mentioned, applying to jobs
that generally have a high volume of applicants helps the study from being detected. Additionally, applying at
different times and changing the names of institutions and organisations listed in the resume can also ensure
that the study is not detected. To strengthen the authenticity email ids, phone numbers, and LinkedIn profiles
were assigned to each ‘candidate’.

The question of other variables is an important one and has been weighed heavily while constructing the two
resumes. There were three concerns here.

Firstly, in entry-level positions with a lower barrier of entry, oftentimes, the first resume submitted is
automatically at an advantage. To combat this, the order of submission was done on a rotation.

Secondly, other factors like differences in the prestige of organisations and socioeconomic class may also
lead to one candidate getting a call-back over another. This was minimised by providing addresses of similar
socixo-economic profiles and choosing institutions to include in the resume in such a way that they are as
evenly matched as possible. Halfway through the study, the resumes were exchanged in order to minimize
the likelihood that the call-backs were due to preference for a resume over the name.

Additionally, the problem that resume studies pose is that they fabricate unrealistically exceptional profiles
of minorities and as such, do not act as an informative measure of discrimination in the job market. In order to
circumvent this problem, the study researched what reasonable qualifications a Muslim woman in her early
20s in an urban setting may have and modelled the resume in a realistic manner. The study considers any
response asking to move to the next round of recruitment as a positive call-back, as a result of this, it cannot
measure who would get the final job, but can tell what biases exist in the first few stages of recruitment. It is
also hard to account for the difference between company biases and individual biases, so our research
explores the overall bias in the labour force. While our analysis of different dimensions of the bias detected is
a starting point, it relies on smaller, non-standard sample sizes, and thus, displays a larger margin for error.

Another limitation of the study is that it does not compare the hiring bias that may come to play between a
Dalit Hindu and Muslim woman, as the name ‘Sharma’ generally connotes an upper-caste Hindu. But, it is
worth noting that, in Attewell and Thorat’s seminal study, they conclude that Muslim men, on average, were
less likely to get call-backs than Dalit Hindu men

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 6

DATA AND
FINDINGS
Over the course of 10 months, approximately 2,000 job applications were sent to 1,000 job postings on job
search sites like LinkedIn and Naukri.com with two dummy profiles, where one was Hindu (Priyanka Sharma)
and the other was Muslim (Habiba Ali). Roles applied for look like content writer, business development
analyst, and social media marketing specialist across industries. The primary analytical goal of the research
was to ascertain a net discrimination percentage by using the number of positive responses to both
candidates. The researchers counted all responses of passage to the next round of recruitment as positive.
Additionally, the researchers also considered instances where companies found ‘Priyanka’ or ‘Habiba’ on
LinkedIn and reached out as a positive response. All statistics found in the course of the research are
≤
statistically significant, with p-values .05.

1. Net Discrimination Rate between Priyanka and Habiba

Table 1

Variable Name Positive Responses to Application

Hindu 208

Muslim 103

Table 1 provides the raw data for the applications and the call-backs. With all positive responses taken into
account, Priyanka got 208 whereas Habiba got only 103. There were no negative responses, rather there were
no responses in non-positive instances. The net discrimination rate gives us a measure of the disparity
between the number of positive responses extended to Priyanka relative to Habiba, displaying the percentage
of cases where Priyanka saw an advantage over Habiba. It is calculated by subtracting the number of cases
where the minority is invited (c) from the cases where the individual belonging to the majority community is
invited (b), divided by the number of cases where at least one candidate has been invited (a). So, if the net
discrimination was 0%, Habiba would get a positive response for every positive response Priyanka received,
whereas if it was 100% Habiba would receive no positive responses for every positive response Priyanka
received.

Net Discrimination Rate = (b - c) / (a) * 100

The net discrimination rate for Indian Muslim women relative to Hindu women then becomes 47.1%,
representing a massive discrepancy between call-backs for Muslim and Hindu women and proving that a
significant hiring bias favouring Hindu women is present across industries. The response rate for Priyanka was
0.208 (20.8%), whereas for Habiba it was 0.103 (10.3%), with the difference in response rates being 0.105
(10.5%). In simpler terms, our research finds that approximately, for every positive response that a Muslim
woman gets, a Hindu woman gets two.

Response Rate

www.reallygreatsite.com 21
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 7
 2. The Nature of Bias

Further findings upon analysing the data qualitatively hint at differing levels of enthusiasm for both
candidates, with ‘Habiba’ receiving only a rare follow-up call, whereas a large proportion of Priyanka’s
responses were in form of call-backs. Similarly, ‘Habiba’ got only 1 positive response from recruiters coming
upon her profile, whereas ‘Priyanka’ received 15 such responses. This displays the difference in proactivity
that recruiters have when they approach the two profiles, with Priyanka being recruited with far more
enthusiasm. These disparities point to more nuances within the biases that exist against Muslim women in
the job market that need further interrogation.

Nature of Responses (Priyanka) Nature of Responses (Habiba)

(a) Role-wise disparity

The research had around 500 data points for creative and content research-oriented roles, the disparity in
response rates for Priyanka and Habiba in this role was 10.7%, with Habiba only getting a positive response
10.3% of the time, and Priyanka getting a positive response 21% of the time. This is in line with the average
difference in response rate.

www.reallygreatsite.com 21
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 8
(a) Industry wise disparity

Over the course of this research, the researchers applied to 90 industries and saw a disparity in the call-back
rates across industries. The researchers had over a hundred data points for Information Technology and
Services (160) and Marketing and Advertising (198) industries each. The difference between the call-back
rates was 11.4% and 9.9% respectively. In both cases, the net discrimination rate was over 50%, with 52.8% for
Information Technology and Services and 57.1% for Marketing and Advertising.

Response Rate (%)

(c) Region-wise Disparity

The job postings were in 20 different states, with a particular concentration in Delhi (198) and Maharashtra
(148), likely due to the concentration of companies in these areas. These locations display a difference of 5%
and 7.2%, respectively. Interestingly, Delhi shows a net discrimination rate of 24.8% only whereas Maharashtra
displays 51.4%, closer to the average, but slightly higher.

Response Rate (%)

www.reallygreatsite.com 21
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 9
 Response Rate across Regions (%)

Different regions also reflect different disparities, according to our data. The researchers saw a concentration
of job applications in the North (400), West (221), and South (144) Zones of India. Here, the difference in
response rates is reasonably uniform across the board, with a 9.1%, 8.6%, and 9.6% difference in response
rates. But, this translates to widely different net discrimination rates. In the case of North India, there is a net
discrimination rate of 39.5%, whereas West and South India show 59.3% and 60%, respectively.

3. Explanation

Statistics show that Muslim women, and women, in general, are poorly represented in the labour force. Such a
disparity demonstrates that there is a direct bias existing against Indian Muslim women in the hiring process.
Muslim women, by virtue of a combination of lack of representation and historical marginalization, are rarely
seen as ‘providers’ and competent workers of the labour force. In their case, a general notion of women being
incapable of working outside the home propounds with the negative stereotypes attached to being Muslim in
India.

Although no quantified evidence exists, the research can comfortably gather that Hindu men occupy most of
the high-level roles in private industries. This is on the basis of the fact that they have the most access to
education and are favoured over men of other communities. These individuals are likely to then set general
tones about who belongs in the workplace and directly participate in recruitment. This disparity can be
explained through subconscious biases, conscious biases, and discriminative policies for hiring. This
explanation is, of course, non-exhaustive, but factors such as these are likely to contribute to the biases
existing in the hiring process.

The difference in the discrimination rates across industries and regions is an interesting phenomenon. For the
two industries in question, the net discrimination rate is close to the average net discrimination rate of 47.1%,
but they also display a slight elevation in discrimination. Similarly, the difference in discrimination rates across
regions also requires further study.

www.reallygreatsite.com 21
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 10
CONCLUSION
This study concludes that a significant hiring bias exists against Muslim women even in instances where they
are equally qualified for the job. Muslim women have the lowest WPR (work participation rate) (Ghosh, 2004);
only 21.6% of Muslim women occupy regular jobs in urban areas as compared to 40.4% and 60.7% for Hindu
and Christian women respectively (Census, 2011). Muslims have the lowest worker-population ratio at 32.6,
whereas Hindus, Christians, Buddhists, Sikhs, and Jains have 41%, 41.9%, 43.1%, 36.3%, and 35.5%,
respectively (Census, 2011). Muslim women tend to work from home to a much larger degree (70%) than
thxae average (51%) (Robinson, 2008). This research proves that a contributing factor to such a stark
disparity in labour market participation is discrimination within the hiring process. Equal access to
opportunities for Muslim women is vital to their social and financial equality in society and tackling biases in
the hiring process is one of the most important mechanisms by which one can level the playing field.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 11

SOLUTIONS AND
RECOMMENDATIONS
While this is an important conclusion to arrive at, it is also a starting point in research that can and should be
conducted with respect to Muslim women’s participation in the labour force and larger issues in the hiring
process. Further research on how the labour market interacts with Muslim women (and other oppressed
classes) and the role of algorithms in supporting biases is crucial to begin formulating solutions to the
disparity noted in this study. For example, research into how this disparity manifests in executive-level
positions or across regions within the country can help lend more clarity to this bias. Similarly, more deep
dives into our analysis can also help facilitate a better understanding of such a bias; this looks like dedicated
research for certain job roles, industries, or regions.

While labour market economists are fairly well-versed with resume studies, management and organisational
psychology scholars have yet to utilise the tool for their research (Adamovic, 2020). Organisational
psychology research can help provide us with a theoretical basis to interrogate the causes and mechanisms
of biases in the job market (Pedulla, 2018). This is crucial to begin devising solutions to the problem. It is worth
noting that different, more objective mechanisms within the hiring process have shown promising results.
Moss and Tilly (1996) note that the usage of universalistic hiring through reliance on exams or tests, reporting
to superiors about applicant pools and hiring outcomes, and formalised collective decision-making can all
contribute to more equitable hiring decisions, and a shift to more objective mechanisms within the hiring
process is recommended.

Blind Hiring Processes: In blind hiring processes, the name and other non-essential signifiers of identity
can be scrubbed from the profile. This can keep signifiers of religion and gender from creeping into the
hiring process.
Work Sample Tests: These tests allow the recruiter to judge the candidates based on their ability to do a
specific task or skills, shifting the focus from other variables.
Panel Recruitment: Having a diverse panel of recruiters allows for a recruitment process that mitigates
individual biases. Including more Muslim women in the hiring process will allow more Muslim women to be
given a fair recruiting process..
Diversity Goals: By setting diversity goals, organisations can prioritize a representative hiring process and
make diversity a priority. It also becomes a metric by which an organization can measure its hiring process
to see if certain groups face an undue advantage.
Further Research: Research into how this disparity manifests in executive-level positions, across regions
within the country, across industries, across job search aggregators, and in more senior roles can help
clarify this bias. Even among Muslim women, there exist further complexities, for example between Hijab-
wearing and non-Hijab-wearing women.
Facilitating Discussions: While structural solutions are crucial, change can also be made on an individual
level. Diversity is often not discussed within workplaces and this allows biases and discrimination to go
unnoticed. Having conversations that increase awareness, building safe spaces, and demanding more
inclusion in the workplace are all effective measures for combating bias.

This is a non-exhaustive list of measures that can be taken to reduce biases. Increased research within
organizations into their diversity profiles and how they can better innovate to accommodate increased
diversity is advised. Inclusivity and diversity in the workplace help both marginalized communities and
businesses. In a 2015 study by McKinsey surveying 366 public companies, companies in the top quartile for
ethnic and racial diversity and gender diversity were found to be 35% and 15% more likely to have financial
returns above their industry mean. Diverse environments also facilitate increased interaction between those
from varied backgrounds, allowing people to interrogate their biases and assumptions. Increasing Muslim
women’s opportunities, in particular, could help reverse India’s falling female labour force participation rate.

www.reallygreatsite.com 21
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 12
CONFLICT OF
INTEREST STATEMENT
This research was conducted by LedBy Foundation, an organization focused on the economic empowerment
of Indian Muslim Women. It was funded by LedBy Foundation and Maulana Azad National Urdu University in
collaboration with Centre for Development of Policy and Practice.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 13

ACKNOWLEDGEMENTS
We thank Amitabh Kundu, Pranav Kothari, and Amirullah Khan for their invaluable inputs and suggestions for
the formulation of this research.

We would also like to express our gratitude to Nadim Hashem, Guil Moussallem, and Sophie Ross for their
insights and critical analysis of the research.

We thank Anam Sherwani, Harshita Kedia and Arshia Bathla for their work on synthesizing the research.

A special thanks to Md Muaz Hyder, for validating data points with the research team.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 14

 REFERENCES
1. Adamovic, M. (2020). Analyzing discrimination in 15. Hunt, V., Layton, D., & Prince, S. (2015). Why diversity
recruitment: A guide and best practices for resume studies. matters. McKinsey & Company. Last accessed at
International Journal Of Selection And Assessment, 28(4), https://www.mckinsey.com/business-functions/people-and-
445-464. Last accessed at organizational-performance/our-insights/why-diversity-matters
https://doi.org/10.1111/ijsa.12298 on June 19, 2022. on June 19, 2022
2. Adesina, Z. and Marocico, O., 2017. Is it easier to get a job 16. Kanter, R. (2010). Men and Women of the Corporation. Basic
if you're Adam or Mohamed?. [online] BBC News. Last Books.
accessed at https://www.bbc.com/news/uk-england- 17. Mascarenhas, A., 2016. Muslim participation lowest among
london-38751307 on June 19, 2022 women in workforce: Report. [online] The Indian Express. Last
3. Allport, G. (1954). The Nature of Prejudice. Addison- accessed at
Wesley. https://indianexpress.com/article/cities/pune/muslim-
4. Baert, Stijn & Rotsaert, Olivier & Verhaest, Dieter & Omey, participation-lowest-among-women-in-workforce-report/ on
Eddy. (2016). Student Employment and Later Labour Market June 19, 2022
Success: No Evidence for Higher Employment Chances: 18. Massey, D. S. (2007). Categorically Unequal: The American
Student Employment and Later Labour Market Success. Stratification System. Russell Sage Foundation. Last accessed
Kyklos. 69. 401-425. 10.1111/kykl.12115. at http://www.jstor.org/stable/10.7758/9781610443807 on
5. Bedi, A., Majilla, T. and Rieger, M., 2018. Gender Norms and June 19, 2022
the Motherhood Penalty: Experimental Evidence from India. 19. Medina-Hernández, E., Fernández-Gómez, M., & Barrera-
[online] Ftp.iza.org. Last accessed at Mellado, I. (2021). Analysis of Time Use Surveys Using CO-
http://ftp.iza.org/dp11360.pdf on June 19, 2022 STATIS: A Multiway Data Analysis of Gender Inequalities in
6. Bertrand, M., & Mullaitnathan, S. (2004). Are Emily and Time Use in Colombia. Sustainability, 13(23), 13073. Last
Greg More Employable than Lakisha and Jamal? A Field accessed at https://doi.org/10.3390/su132313073 on June 19,
Experiment on Labor Market Discrimination. The American 2022
Economic Review, 94(4), 991-1013. 20. Moss, P., & Tilly, C. (1996). “Soft” Skills and Race: Work
7. Carlsson, M., Dahl, G., & Rooth, D. (2012). The Effect of And Occupations, 23(3), 252-276. Last accessed at
Schooling on Cognitive Skills. SSRN Electronic Journal. Last https://doi.org/10.1177/0730888496023003002 on June 19,
accessed at https://doi.org/10.2139/ssrn.2164642 on June 2022
19, 2022 21. Pager, D. (2007). The Use of Field Experiments for Studies
8. 2011 Census of India. Office of the Registrar General & of Employment Discrimination: Contributions, Critiques, and
Census Commissioner, India. India. [Web Archive] Retrieved Directions for the Future. The ANNALS Of The American
from the Library of Congress. Last accessed at Academy Of Political And Social Science, 609(1), 104-133. Last
https://www.loc.gov/item/lcwaN0017959/ on Jun 19, 2022 accessed at https://doi.org/10.1177/0002716206294796 on
9. Collins, R. (1979). The Credential Society. New York: June 19, 2022
Academic Press. 22. Pedulla, D. (2018). How Race and Unemployment Shape
10. Darity, W. (2003). Employment Discrimination, Segregation, Labor Market Opportunities: Additive, Amplified, or Muted
and Health. American Journal Of Public Health, 93(2), 226- Effects?. Social Forces, 96(4), 1477-1506. Last accessed at
231. Last accessed at https://doi.org/10.1093/sf/soy002 on June 19, 2022
https://doi.org/10.2105/ajph.93.2.226 on June 19, 2022 23. Riach, P., & Rich, J. (2002). Field Experiments of
11. Darity, W., & Deshpande, A. (2000). Intergroup Economic Discrimination in the Market Place. The Economic Journal,
Inequality Across Countries: An Introductory Essay. Review 112(483), F480-F518. Last accessed at
Of Social Economy, 58(3), 273-276. Last accessed at https://doi.org/10.1111/1468-0297.00080 on June 19, 2022
https://doi.org/10.1080/00346760050132328 on June 19, 24. Thorat, S., & Attewell, P. (2007). The Legacy of Social
2022 Exclusion: A Correspondence Study of Job Discrimination in
12. Duflo, E. (2012). Women Empowerment and Economic India. Economic and Political Weekly, 42(41), 4141–4145. Last
Development. Journal Of Economic Literature, 50(4), 1051- accessed at http://www.jstor.org/stable/40276548 on June
1079. Last accessed at 19, 2022
https://doi.org/10.1257/jel.50.4.1051on June 19, 2022 25. Weber, M. (1968). Economy and society. University of
13. Eriksson, S., & Rooth, D.-O. (2014). Do Employers Use California Press.
Unemployment as a Sorting Criterion When Hiring? 26. Woetzel, J., Madgavkar, A., Sneader, K., Tonby, O., Lin, D., &
Evidence from a Field Experiment. The American Economic Lyndon, J. et al. (2018). The power of parity: Advancing
Review, 104(3), 1014–1039. Last accessed at women’s equality in Asia Pacific. McKinsey & Company. Last
http://www.jstor.org/stable/42920727 on June 19, 2022 accessed at https://www.mckinsey.com/featured-
14. Gupta, D. (1992). Social stratification. Oxford University insights/gender-equality/the-power-of-parity-advancing-
Press. womens-equality-in-asia-pacific on June 19, 2022

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 15

APPENDIX
Resumes: First set of two resumes were used for the first 500 applications. The second set of two resumes
were used for the second set of 500 applications.

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 16

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 17
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 18
HIRING BIAS REPORT | 2022 LEDBY FOUNDATION | PAGE 19
KEY CONTACTS
Dr. Ruha Shadab (ruha@ledby.org)
Deepanjali Lahiri (deepanjali@ledby.org)
Vanshika Sharan (vanshikasharan@gmail.com)

Reporters and news media

Please direct requests to

Deepanjali Lahiri
deepanjali@ledby.org
Tel: +91 9967 954121

For more information, visit https://www.ledby.org/hiringbias

Designed by
Arshia Bathla

HIRING BIAS REPORT | 2022 LEDBY FOUNDATION