In the months following 9/11, considerable evidence suggests that anti-Islamic sentiment
rose in the US and resulted in an increase in hate crimes. Anti-Islamic sentiment
also likely extended to more subtle impacts such as work-place discrimination. Rigorously
documenting these patterns for a representative sample of Muslims is not straightforward.
Few surveys contain detailed information on religion, and those that do don’t have
sufficient numbers of Muslims for most analytical purposes. Moreover, the potentially
negative implications of 9/11 likely extend beyond practicing Muslims, to include
those likely to be perceived as Muslims.
I am going to examine changes over time in a set of outcomes for Arab and Iraqi-Pakistani-Afghani
(IPA) men aged 25-40 with the idea that if abrupt changes shortly after 9/11 are concentrated
among those groups, it is suggestive that 9/11 caused changes that differentially
affected those likely to be viewed as Muslims. In other words, if individuals of ancestries
suggesting they are Muslim experienced substantial discrimination after 9/11, we may
be able to see evidence of it in the form of changes for these groups alone.
This study uses observations from the American Community Survey (ACS) for years 2000-2007.
I downloaded the ACS data from the University of Minnesota Population Center integrated
public use micro-data series. As all answers in the ACS are recorded as outcomes in
the “previous year,” the data actually reflects the years 1999-2006.
There are 4 treatment groups relative to 3 control groups utilized in this study.
I focus on males that are aged 25-40. The four treatment groups are: (1) Immigrant
Arabs (2) US-born Arabs (3) Immigrant IPA (4) US-born IPA. There are three control
groups are: (1) European Whites (2) Immigrant East Asian (3) US born East Asian.
This study utilizes difference-in-difference analysis through linear regressions to
assess whether the time pattern in outcomes changed significantly and abruptly after
9/11 for prime-age men of Arab or IPA ancestry relative to those of European White
ancestry. I also compare Arab and IPA men with those of East Asian ancestry in order
to see if different patterns arise when comparing the treatment groups against another
minority group of similar education and professional background.
I use 6 outcome variables:
• Works full time- a dichotomous variable indicating whether the respondent has a
full time job.
• Any work- a dichotomous variable indicating whether the respondent has worked at
least one hour.
• Employment- a dichotomous variable indicating whether the respondent is employed
in the labor force.
• Annual hours worked- the total number of hours a respondent worked.
• Hourly earnings-constructed as total income divided by total annual hours work which
is then logged.
• Cognitive difficulty- a dichotomous variable which represents whether respondents
have physical, mental, or emotional difficulties that have impaired their abilities
for longer than 6 months.
Results & Discussion
In the US as a whole, evidence of perceived discrimination possibly due to 9/11 is
slim. Possible 9/11 effects may be seen in the lack of increase in the rate of any
work for immigrant Arab men and US-born IPA men shortly after 9/11. Additionally,
a possible 9/11 effect may explain a short-term decrease in hourly earnings for US-born
Arab men and immigrant Arab men who work full time.
A more concentrated impact is seen for the treatment groups in the state analysis.
A lack of increase in the rate of working full time shortly after 9/11 for immigrant
IPA men in DC-Maryland-Virginia (DMV) may be a possible 9/11 effect. A similar lack
of increase in employment after 9/11 for US-born Arabs in Michigan may also be the
result of a 9/11 effect. Immigrant Arab men experience a decrease in earnings per
hour shortly after 9/11 in both DMV and Michigan, possibly related to an immediate
impact from a 9/11 effect. The treatment groups in New York experience the most changes
across the outcomes that may be a result of a 9/11 effect. The rate of any work is
lower for immigrant Arab men after 9/11 in New York. Additionally, for US-born Arabs
who work full time in New York the only year in which they do not experience an increase
in earnings is 2003.
Changes in employment patterns that may be a result of a 9/11 effect is also observable
in specific industries. The rate of being employed for immigrant IPA men is lower
in the manufacturing and professional/scientific industry shortly after 9/11. The
rate of any work for immigrant IPA men in the retail trade industry remains unchanged
in 2003 while rising in every other year. US-born Arab men earn less shortly after
9/11 in the retail trade industry for all workers and full time workers. Immigrant
and US-born Arab men who work full time earn less shortly after 9/11 in both the manufacturing
and retail trade industry.
Changes in cognitive difficulty are limited to immigrant IPA men in the manufacturing
industry and US-born IPA men in the professional/scientific industry. The increased
rate of cognitive difficulty for both groups may be the result of a 9/11 effect due
to their immediate short-term impact. The 5.4% increase in the rate of cognitive difficulty
in 2003 for immigrant IPA men in manufacturing shows a possible 9/11 effect beyond
traditional employment patterns. A similar possible 9/11 effect on cognitive difficulty
for US-born IPA men in professional/scientific industry is even more telling. US-born
IPA men do not experience any significant difference in any outcome compared to European
White men other than cognitive difficulty. But, the 36.1% increase in the rate of
cognitive difficulty in 2003 and 32.2% increase in 2004 shows that even high skilled
jobs may be impacted from a possible 9/11 effect.