|
A UNIVERSITY EMP=
LOYER
DISCRIMINATION COMPLIANCE ANALYSIS/ALERT SYSTEM AND THE CONVERGENCE OF
FORENSIC EXPERTS IN ACCOUNTING AND C=
OMPUTING |
|
|
|
Avi Rushinek, Ph=
.D. Sara Rushinek, P=
h.D. Joel Stutz, Ph.D=
. Please Send
Correspondence To: Sara Rushinek University of Mi=
ami CIS Dept. 421 Je=
nkins Coral Gables, Fl=
33124 305.666.7890 s.rushinek@miami=
.edu |
|
|
|
|
AN EMPLOYER DISCRIMINATION COMPLIANCE ANALYSIS/ALERT SYSTEM AND THE CONVERGENCE OF FORENSIC EXPERTS IN
ACCOUNTING AND COMPUTING
Abstract
This
study develops a discrimination compliance analysis system that decides whe=
ther
or not organizations comply with anti-discrimination regulations (United St=
ates
Equal Employment Opportunity Commission, US EEOC). The US EEOC states that
organizations may not discriminate against employees on the basis of countr=
y of
origin (national origin). The trouble is that such organizations may not be
aware of discrimination in time. Thus, we develop a web server alert system
that combines the knowledge and skills of forensic experts in accounting and
computing to provide early warnings. Using Multiple Cross-Sectional Time-Se=
ries
Linear Regression Analysis the system can determine whether there is
statistically significant discrimination based on national origin, sex, rac=
e,
religion, and/or disabilities. It cannot determine whether this discriminat=
ion
is legal or not; it only tells us that such statistically significant
discrimination exists. Moreover, this system will help calculate the damages
and recoveries to settle such cases expediently.
AN EM= PLOYER DISCRIMINATION COMPLIANCE ANALYSIS/ALERT SYSTEM AND THE CONVERGENCE OF FORE= NSIC EXPERTS IN
ACCOUNTING AND COMPUTING
I. Introduction
This study develops an Alert Control methodology for a
Forensic Accounting & Computing Expert System (FACES). This system is
intended to assist employers in avoiding lawsuits resulting from violations=
of
provisions of the Equal Employment Opportunity Commission (EEOC) or the
Americans with Disabilities Act (ADA). As such, this system would produce a=
lerts
that will warn management of red flags that may reveal the existence of ill=
egal
discrimination. Management can then respond by correcting the problem as so=
on
as possible, if such a problem exists, thereby avoiding possible lawsuits. =
On
the other hand, if management chooses not to correct the problem, this may
result in plaintiffs pursuing a lawsuit that is motivated by an employment =
law
firm willing to fund and underwrite such a lawsuit. This study develops a w=
eb
server that collects relevant EEOC/ADA and job offer data. Then, applying a
model using Multiple Cross-Sectional Time-Series Linear Regression Analysis=
the
system can determine whether there is statistically significant discriminat=
ion
based on national origin, sex, race, religion, and/or disabilities. It cann=
ot
determine whether this discrimination is legal or not; it only tells us that
such statistically significant discrimination exists. The question of legal=
ity
is beyond the scope of the current study, even if it may be a very importan=
t question
for future research. It should be noted that although this system is primar=
ily
concerned with violations of EEOC and
The design, development, deployment, and maintenance of a
FACES system go beyond the capabilities of either the typical forensic
accounting or forensic computing expert system can handle alone. The Comput=
er
Litigation Support Affirmative Class Action Suit Alert Control system,
primarily a preventive control that used to be part of the traditional inte=
rnal
controls of an accounting system, is unable to handle the complexities of
current Internet-based accounting systems. The deployment of current
technologies such as an Internet Web Server and automatic continuous publis=
hing
of Multiple Cross-Sectional Time-Series Linear Regression Analysis to the W=
WW
(World Wide Web), required by FACES, takes a bit more than the typical fore=
nsic
accounting system can deliver by itself. That is where the collaboration and
convergence between forensic accounting and computing occur. The collaborat=
ion
between forensic accounting and computing also extends to the formation of
expert witness testimony teams in the event of a lawsuit.
The sample data use=
d in
this study deals with discrimination lawsuits against universities based on
EEOC/ADA violations pertaining to university professors’ origin,
disabilities, religion, and/or sex. Academics in areas such as MIS (Managem=
ent
Information Systems), CIS (Computer Information Systems), IS (Information Systems)=
, CS
(Computer Science), Computer Engineering, Computer Auditing, and other comp=
uter
related fields in particular, may=
be
more sensitive to some EEOC/ADA violations than other professionals. These areas are un=
ique
in some ways. This may be due to =
the
rapid technological changes that these professionals have to deal with on a
daily basis, requiring an extraordinary continuing professional education, =
as
well as the related cost structures. Denial of such education and exposure =
to
the advances of technology can easily ruin their career growth. Another rea=
son
for special sensitivity to such EEOC/ADA violations may be related to the
relatively large number of foreign professionals with different national
origins and a larger number of individuals with disabilities. Being a forei=
gner
and/or a person with disabilities may further restrict their professional
mobility and opportunities. Therefore, illegal discrimination based on EEOC=
/ADA
and national origin may have more sever effects on these professionals comp=
ared
to other professions. That is especially true if such discrimination depriv=
es
them of desperately needed continuing professional education and practice in
the application of new technologies.
The FACES system runs as an Internet web server providing
subscription services and uses calculation models of multiple linear regres=
sion
analysis. Frequently, any litigation and their related calculations are very
complicated, especially whenever tenured professors are involved, due to the
long periods of time and the inability of a university to easily fire these
professors.
II. Equal Employment Opportunity Commi=
ssion
(EEOC) & Americans with Disabilities Act (
The U.S. Equal Employment Opportunity Commission (EEOC) crea=
ted
a variety of legal issues. These issues include providing equal employment
opportunities to protected categories of employees based on protected
categories of national origin, sex, race, religion, and/or disabilities. The
EEOC issued this report on the occasion of the 10th Anniversary =
of
the signing of the Americans with Disabilities Act (ADA). Title I of the
As expressed by Ida L. Castro, Chairwoman of the EE= OC:
The EEOC has taken an active and forceful =
role
in removing barriers and increasing opportunities for people with disabilit=
ies
in the workplace. More than 21 percent of the Exec s caseload is comprised =
of
charges filed under the
Too
many barriers remain for people with disabilities in finding and keeping jo=
bs,
getting promotions, obtaining reasonable accommodations, and in other aspec=
ts
of the employment process. Qualified individuals with disabilities, with or
without reasonable accommodation, are entitled to the same employment
opportunities available to people without disabilities. This is the purpose=
and
the spirit of the
A steady income is of critical importance,=
but
it is only one of the many advantages of employment. A decent job can enhan=
ce
self-worth, provide educational opportunities and skills training, give one=
's
life structure and purpose, increase social contacts, and offer important
fringe benefits such as health insurance, retirement pensions, travel
opportunities, and paid vacation time. Lack of employment can lead to pover=
ty,
stagnation, loss of self-esteem, and isolation.
From the
These numbers help the Forensic Accountant calculate the
mathematical expectation of collecting damages. As an example, if faculty
members’ disability claims total damages of $2,000,000, and 15% of the
cases are ruled in favor of the plaintiff, the FA should calculate the cash
inflows as at least $300,000 =3D $2,000,000*15%.
III. The Role of Forensic Accounting
The Forensic Accoun=
tant
(FA) can verify that discrimination exists and is statistically and financi=
ally
significant. The
FA can calculate the damage that results from the lack of a steady income d=
ue
to discrimination and show the statistical significance of the drop in inco=
me.
A job that is not decent due to discrimination can hurt a person’s
self-worth, especially when it denies educational opportunities and skills
training, deprives one of one's life structure and purpose, decreases social
contacts, denies important fringe benefits, etc. In the case of university
faculty in can also include lose of conference travel opportunities, gradua=
te
student assistance, choice courses and desirable schedule, computer upgrades
and other resources, etc. These discriminatory policies lead to stagnation,
loss of self-esteem, and isolation, which lead to pain and suffering, health
deterioration, and other damages that the FA can calculate. Economists have
traditionally calculated these hedonic damages, but due to cost cuts,
accountants are forced to deal with these issues.
In
our sample case, a supervisor caused undue stress to a Professor with a
disability by assigning him a 5-day teaching schedule (Monday through Frida=
y),
spread throughout the day (classes at 8:00=
a.m.,
12:00 p.m., and 6:00 p.m.), in a remote building 15 minutes from his office=
. At
the same time other faculty members, without disabilities, received a 2-day
schedule with back-to-back classes in the same building as their office.
In
addition, a forensic accounting expert can assist in conducting a forensic
audit that will evaluate an organization’s internal controls.2 <=
/sup>Based
on the audit, the FA can attest to the existence and effectiveness of the
controls to enforce the EEOC requirements. Likewise, a FA can help forecast=
the
cost associated with EEOC compliance or the cost of not complying.3
IV. The Role of Forensic Computing
Litigation has a host of problems, the main ones being the
escalating costs of filing a lawsuit and the enormous costs of concluding t=
hem
successfully.16 A Forensic Accountant (FA) can help by applying
capital budgeting to decide how cost effective is the lawsuit, skills that a
forensic computing expert and the legal team may not have; for that reason,
many lawsuits are settled or not litigated at all. To deal with the problem=
of
rising costs, a Forensic Computing Expert (FCE) expert can help automate and
streamline some of the activities, reducing the costs and speeding the proc=
ess
of preparation. For example, instead of filing lots of small lawsuits, the
plaintiffs can file a class action lawsuit that is more cost effective and =
can
be more lucrative for the law firm. The computer litigation support of the =
FCE
can provide the computing infrastructure and web services to reduce the cos=
ts
of such class actions to a level that makes them cost effective. Lawsuits t=
hat
are not cost effective by themselves may become cost effective as a class
action, provided that a forensic computing expert can streamline them, as we
will describe in this study.
The “Hard Data” of a large number of professiona=
ls
(approximately 650 in the AIS/ISWORLD salary survey data) is more objective
than the subjective discrimination claims of any individual or small group =
of
individuals. For these reasons law firms will be more competitive with each
other and more eager to represent an entire class of computer professionals.
Likewise, law firms will be more likely to underwrite and fund such class
action suits against deep pocket employers who can afford to pay significant
damages without going under.
In the early days of forensic accounting the calculations we=
re
relatively simple and the companies involved were relatively small. Today, =
the
complexity has grown dramatically and a computer is essential for litigation
support, especially in large cases such as: Enron Corporation; Global Cross=
ing
Corporation; World-com Corporation; IMClone and Martha Stewart Living (Insi=
der
Trading); the
What are some of the issues that would require forensic
computing expertise? Enron
Corporation provides an excellent example of the invaluable services of a
forensic computing expert. Due to the massive document shredding many of the
normal incriminating hard copy accounting records might have been destroyed.
What may be necessary to reconstruct those accounting records are not only a
team of forensic accountants but also some forensic computing experts who,
through the reconstruction of complicated computer systems, such as futures
contracts trading systems, can regenerate records that are similar those th=
at
had been shredded.
Another example is the current American war against terrorism
that is trying to track the world wide web of terrorist communications and
transactions supporting organizations such as Islamic Jihad, Al-Qaeda,
Hezbollah, and others on the
Additional issues that require FCE involvement are adherence=
to
privacy and security laws. It is important that data is properly encrypted,
secured, and audit trails maintained; otherwise, multiple laws can be easil=
y violated.18
For example, disclosure of certain records and information (e.g.,
medical) to the public may ruin an individual’s reputation. This can
border on computer crime, if a high ranking administrator gets into the med=
ical
records of an individual, without proper authority, falsifies that record a=
nd
misuses it to bar the individual from work on false pretence. Therefore, such records should be
protected with proper password procedures and not available for public rele=
ase.
An FCE can verify that proper data encryption and access controls exist and=
are
effectively enforced. The legal issues are beyond the scope of this study b=
ut
it is clear that to evaluate the computer controls an FCE may be needed.
V. FACES – A Software Discrimin=
ation
Compliance Alert System
Forensic accounting and auditing of a computerized accounti=
ng
information system can help built an infrastructure and controls that will
enforce EEOC compliance.4 Such controls will ensure that promoti=
ons,
raises, rank, salary, and conditions of employment do not violate EEOC
requirements. It will ensure that performance and compensations are properly
correlated and correspond to industry benchmarks as reflected by the indust=
ry
averages. Computer output controls will ensure that if EEOC compliance is
violated, the proper personnel, such as the legal counsel, will be advised
about the problems so that remedial action can be taken. A database system
should track the compliance issues, the remedial action taken or not taken,=
and
an explanation of the outcome.
EEOC co= mpliance software should be included in generalized audit software, especially for l= egal audit of large organizations where EEOC compliance can easily be overlooked. This should be part of the Payroll and Personnel Software audit. An EEOC compliance audit will not only reduce the risk of the organization, but also most likely raise moral and computer user satisfaction. To ensure complianc= e, the forensic accounting & computing team, should ensure that proper bac= kup and recovery procedures of the audit report and of the web server exist and= are working effectively. Such systems are beyond the scope of forensic accounti= ng alone since they are dealing with expert systems and artificial intelligenc= e, which usually will require the expertise of a forensic computing expert, especially for the design and development phase.
The EEOC establishes standards of equal employment
opportunities. Our system (FACES) applies Accounting
Variance Analysis Theory and sends alerts to the existence of any red
flags that show deviations from those standards. As Rushinek (1983)12 describes in "Control Aspects of
Standard Costing: Variance
Analysis, Inflation Adjustment, the Learning and Their Computer
Applications," it is dealing with deviations from standards. How can we
evaluate these deviations and take corrective action? In "Control thro=
ugh
Standard Costs and Variance Analysis for Performance Evaluation," Rush=
inek
& Rushinek (1990)13 apply Variance Analysis to the improveme=
nt
of profitability.
Using Statistical ANOVA Theory Applied to E= EOC standards, FACES Regression Analysis tells us whether or not the differences between United States (US) nationals and employees from other national orig= ins are statistically significant. This is the initial stage of the forensic investigation. If no differences exist or if the differences that do exist are statistically insignificant, there is no point pursuing the subsequent secondary stage, the accounting analysis of variance.
This accounting ANOVA will tell us, as in our case s=
tudy
of a hypothetical
Using t=
he
Accounting Variance Analysis Theory Applied to EEOC standards, FACES tells =
us
whether or not the differences between the
It should be clear that a Web Server based survey is much mo=
re
cost effective than the traditional soft (non-quantitative) way of alerting=
one
to class action opportunities. The web server software is free, including
Microsoft Internet Information Server (Bundled with Windows XP, Professiona=
l,
Server, and Advanced Server) or Apache (Free download from apache.org). Likewise, the operating system is
usually bundled with the hardware or is a free download (Linux). The standa=
rd
DSL (Digital Subscriber Line) bandwidth that a text based web server consum=
es
is very inexpensive (about $30-$60 a month).
VI. Data Collection
A web server on the WWW (World Wide Web) provides a survey t=
hat
collects data for the regression model. Since the server performs all of the
activities automatically, it can update information very efficiently,
effectively, & inexpensively in real-time. Appendix A describes a modif=
ied
2002 Computer Professional Offer & Affirmative Action Survey Form. The
survey form lets faculty members respond anonymously and assures them that
their identity and privacy is protected, thus encouraging people to
participate. The form can be used to enter new data or to update/correct da=
ta
that was previously entered. The original survey form can be found at: http://www.pitt.edu=
/~galletta/salsurvanon.html.
Additional items were added to the original form in order to
obtain data required by FACES to provide alerts based on discrimination.
Therefore, the biggest difference between our approach and existing approac=
hes
it that we have a forensic focus in addition to the simple salary focus. Us=
ing
our approach we can achieve all the other objectives but also provide some
guidelines for class action based on discrimination of protected categories
such as national origin. This can be used not only for expert witness testi=
mony
and litigation support services but also for a red flag of potential
discrimination discovery and warning. Thus, we can provide a built-in inter=
nal
control for enforcing affirmative action and preventing illegal discriminat=
ion,
nipping it in the bud before it is litigated.
In order to discover candidates that may result in class act=
ion
suits one needs to analyze a large sample size. For the analysis in Tables =
1,
2, and 3 we have used a total of 608 observations from the AIS/ISWorld surv=
ey
from the following three academic years: (2001-2002), (2000-2001), and
(1999-2000). Each observation represents the survey response of a prospecti=
ve MIS
faculty member.
Table 1 shows the Summary Output of the multivariate regress=
ion
statistics. It shows a low Multiple R of 0.734211535 and an
Table 2 shows the ANOVA (Analysis Of Variance) values for F =
and
Significance F of 43.19880222 and 2.33182E-88 respectively, which is
statistically significant at the .05 level of statistical significance.
Rushinek, A. and Rushinek, S. have used a similar analysis in a previous pa=
per
for a case study on the oil and gas industry.26 The statistical
significance of this ANOVA means that we can reject the null hypothesis that
the R-Square is equal to zero in favor of the alternate hypothesis that the
R-Square is greater than zero. Likewise, it implies that the independent
variables can explain the variance in the dependent variable at the level of
the value of the R-Square. Furthermore, the adjusted R-Square can explain t=
he
same pattern in the population of other universities and cases beyond the
current sample. Thus, if we encounter another case of significant deviation
from the standard, in universities that resemble this sample, it may indica=
te a
case of discrimination on the basis of national origin of the faculty membe=
r.
Table 3 lists the multivariate regression independent variab=
le
names and their statistics, including Coefficients, Standard Error, t Stat,
P-value, Lower 95%, and Upper 95%. The list of variables includes the count=
ry
of the school making the offer, which for our purposes is the most importan=
t.
VII. Data Analysis
Like every other profession, the EEOC provides protection to
computer related professionals and given their unique knowledge the Forensic
Computing Expert (FCE) can provide valuable help in such cases. Discriminat=
ion
that is based on national origin may be more common in areas such as comput=
er
professionals since the
Faculty with discrimination complaints against their univers=
ity
would usually first report them to a committee of the faculty senate. If the
university does not provide the proper remedies faculty members tend to file
the complaints with the EEOC. If the EEOC cannot file a lawsuit or provide =
the
proper remedies a faculty member can file a lawsuit. Such cases typically
include (but are not limited) to the following complaints: denial of raises,
promotions, tenure, summer research support, sabbatical, graduate assistanc=
e,
additional courses for extra income, personal computer upgrades, travel
expenses, choice office, desirable teaching schedule times, reduced teachin=
g load,
advanced graduate course assignments, etc.=
Limiting faculty members to teaching only introductory and elementary
courses thus denying them the opportunity to teach advanced graduate courses
and form close interactions with graduate students restricts them from
advancing their domain knowledge and sharing/enhancing their research proje=
cts.
The
In order to illustrate the potential=
of
the FACES system we will use some of the data values in Tables 1, 2, and 3 =
and
apply several assumptions. We will assume that the tables represent the val=
ues
from a univariate regression with zero intercept. We will also assume that
Country is the sole independent variable and indicates the country of natio=
nal
origin of the faculty respondent, equal to 1 for a US national and 0 for a
foreign national (non US). Salary remains as the dependent variable. Then, =
we
can use the coefficient of Country from Table 3 to estimate the normal diff=
erence
in salary between an MIS faculty member of foreign origin and MIS faculty w=
ith
Since
the $10,400 average salary difference can be due to a number of factors, we=
need
to determine what percentage of the difference can be attributed to a facul=
ty member’s
country of origin. The R-Square value in Table 1 provides this value. If th=
ere
were no discrimination based on national origin, the R-Square of 54% tells =
us
that the coefficient of Country should explain about 54% of the salary
difference between US and foreign nationals. In other words, the salary
difference should equal the rounded Country coefficient times the R-Square =
($10,400
* 0.54), or about $5,600. Therefore, on average, a US MIS faculty member sh=
ould
be paid about $5,600 more than a foreign national MIS faculty member, all o=
ther
things being equal. The combination of high t=
Stat
value (4.288054262) and low P-value (0.000021) in Table 3 rejects the null
hypothesis that national origin causes no disparity in salary and accepts t=
he
alternative hypothesis at the .05 level of statistical significance. The
alternative hypothesis says that national origin is a statistically signifi=
cant
factor in determining the salary of MIS faculty. If there is no other appar=
ent
reason for this disparity, such as that computer faculty members with forei=
gn
national origin have inferior research, than it may be a result of illegal
discrimination on the basis of national origin.
Having
established the expected faculty salary difference based on country of nati=
onal
origin, we created a hypothetical institution,
The res=
ults
show how US nationals compare to foreign nationals at
The res=
ults
further show that
This st= udy assumes that significant numerical disparities are the result of discrimina= tion and not some other unknown reason. In other words, the system first determi= nes if any discrimination exists based on various categories such as national origin. If the differences in values are not significant, for example if US= and Foreign Nationals salaries are about the same, the system will accept that = as non-discrimination and not look any further. A computer web server and a FA= CES type system can independently make such decisions. It can check the statist= ical significance of national origin as an independent variable in the salary mo= del and it will tell you if there is any basis for looking further.
The summary data in Appendix B shows the average discriminat=
ion
damages for foreign nationals (plaintiffs) versus US nationals at imaginary=
The problem is that the $517,113 total potential revenue may=
not
be sufficient for a law firm to consider funding a lawsuit on a contingency
basis and the cost of such a lawsuit could be too much for most employees to
consider on top of the mental and emotional stress that such a lawsuit would
involve. If a law firm could turn this case into a class action suit involv=
ing
several institutions with several employees each, then the substantial incr=
ease
in revenue would make the case more attractive to the law firm, especially
given the high profile. The second section of the table in Appendix C shows=
the
possible outcome of a class action lawsuit involving twenty institutions wi=
th
an average of three plaintiffs at each one, resulting in a total of $31,026=
,780
expected revenues/costs to the plaintiffs/defendants.
&=
nbsp; The
last section of the table in Appendix C computes the possible costs to the =
law
firm of putting a web server on the WWW (World Wide Web) for up to ten year=
s of
deliberation. A professional association, such as ACM (Association for
Computing Machinery), could typically provide such web server services to i=
ts
members and the community. The estimated costs of such a service would incl=
ude
a one-time fee for server setup, configuration, and programming of $500; a
yearly charge for hosting of $500, totaling $5,000 over ten years; plus an
interest fee on hosting of $250; resulting in a total of $5,750.
Subtracting the web server cost of $5,750 from the $31,026,7=
80
expected revenue computed above yields an expected net class action cash in=
flow
(excluding legal fees) of $31,021,030. This positive and high Return on
Investment (ROI) may lead to an overall reduction in discrimination lawsuit=
s; a
system like FACES and the web server would thus prove to be an excellent
deterrent by alerting employers to the possibility of discrimination lawsui=
ts
and allowing them to take actions to avoid them. This may be the most impor=
tant
contribution of the FACES effort. It could result in making the job market =
more
competitive, opening up opportunities for previously discriminated employee=
s,
and insuring a level playing field for compensation packages.
Finally, it is important to insure the reliability of
information provided by the web server software system so that the courts c=
an
rely on it. As such, it may be subject to the cost of effective audits and
e-commerce security controls even if it is free of charge. Likewise, general
computer security is also very important since we do not want the analysis =
to
be biased by unauthorized access to the web server and the software.
VIII. Summary, Conclu=
sions,
& Implications
This study developed an employer
discrimination compliance analysis system that decides whether or not
organizations comply with anti-discrimination regulations such as those of =
the
United States Equal Employment Opportunity Commission (US EEOC). The US EEOC
states that organizations may not discriminate against employees on the bas=
is
of country of origin (national origin). This is especially important for
organizations of higher education that get US federal grants. The problem is
that such organizations may not be aware of discrimination and fail to take
action to “nip it in the bud” before it is too late and a varie=
ty
of sanction are applied including grant denial, class action litigations, b=
ad
press, and excessive legal expenses. Thus, we develop an alert system that =
will
use the convergence of the knowledge and skills of forensic experts in
accounting and computing to provide internal controls to prevent the problem
early on. It
demonstrates how an alert system providing early problem detection and
subsequent enforcement can be automated to raise the level of compliance wi=
th
the law at a lower cost. Likewise, it demonstrated how such efforts can be =
used
effectively for expert witness testimony and computer litigation support in=
EEOC/ADA
cases. Moreover,
this system will also help calculate the damages and recoveries to settle s=
uch
cases expediently.
This study concludes that a web server can help enforce
computer professional EEOC affirmative action. Computer professionals, and =
MIS academic
faculty in particular, may be more sensitive to some EEOC violations than o=
ther
professionals; therefore, such protection from discrimination may also be m=
ore
important for them. This may be due to the rapid technological changes that
these professionals have to deal with on a daily basis, requiring an
extraordinary continuing professional education. Denial of such education a=
nd
exposure to the advances of technology can easily ruin their career growth.
Another reason for special sensitivity to such EEOC violations may be relat=
ed
to the relatively large number of MIS faculty of non-US national origin emp=
loyed
at US colleges and universities. Therefore, illegal discrimination based on
national origin may have a more severe effect on these professionals compar=
ed
to others.
The system, Forensic Accounting & Computing Expert
System (FACES), developed in this study uses an Alert Control methodology to
produce alerts that will warn management of conditions that may reveal the
existence of illegal discrimination. Management can respond by correcting t=
he
problem as soon as possible, if such problem(s) exist. On the other hand, if
management chooses not to correct the problem(s) then this same system, FAC=
ES,
could inspire the aggrieved parties to pursue a lawsuit and motivate some
employment law firms to fund and underwrite such a lawsuit. This study deve=
loped
a web server system that collects EEOC and job offer data and applies a mod=
el using
Multiple Cross-Sectional Time-Series Linear Regression Analysis. This analy=
sis
can tell whether there is a statistically significant discrimination based =
on National
Origin, Sex, Religion, and Disabilities. It cannot determine if the
discrimination is legal or not; it only tells us that such statistically
significant discrimination exists. Whether it is legal or not is another
question? The answer to such a question is beyond the scope of the current
study, even if it may be a very important question for future research.
In
the future, as the technology improves and the business environment relies =
more
and more on computing, forensic computing will become more and more importa=
nt
in combination with forensic accounting. Forensic work will require more of=
a
team effort since the legal and the accounting professionals will need more
help from forensic computer professionals, as this case demonstrates.
TABLE 1: SUMMARY OUTPUT |
||
|
Regression Statistics |
|
|
|
Multiple R |
0.734211535 |
|
|
|
0.539066578 |
|
|
Adjusted |
0.526587839 |
|
|
Standard Error |
13786.40764 |
|
|
Observations |
608 |
|
|
TABLE 2: |
||||||
|
&nbs=
p; |
df |
SS |
MS |
F |
Sign=
ificance
F |
|
|
Regression |
16 |
1.31=
369E+11 |
8210=
581887 |
43.1=
988 |
2.33=
182E-88 |
|
|
Residual |
591<= o:p> |
1.12=
328E+11 |
1900=
65035.7 |
|
|
|
|
Total |
607<= o:p> |
2.43=
698E+11 |
|
|
|
|
|
TABLE 3: INDEPENDENT
VARIABLE NAMES & STATISTICAL ANALYSIS OF SIGNIFICANCE |
||||||
|
&nbs=
p; |
Coefficients |
Standard Error |
t Stat |
P-va=
lue |
Lower
95% |
Upper
95% |
|
Intercept |
6354.134211 |
6.666355409 |
6.02E-11 |
29879.4889 |
54838.345 |
|
|
Revealed? |
-506.6486014 |
1214.087748 |
-0.417308059 |
0.676605 |
-2891.099452 |
1877.80225 |
|
Teaching |
1031.941469 |
143.0969274 |
7.211485863 |
1.7E-12 |
750.9011649 |
1312.98177 |
|
Education |
10056.93747 |
2346.065026 |
4.286725798 |
2.12E-05 |
5449.29955 |
14664.5754 |
|
$Summer |
0.280140683 |
0.051599666 |
5.429118197 |
8.28E-08 |
0.178799683 |
0.38148168 |
|
Summer Years |
68.94572711 |
427.0313533 |
0.161453548 |
0.871791 |
-769.7377 |
907.629154 |
|
Research |
0.704906328 |
0.1595413 |
4.418331368 |
1.18E-05 |
0.391569514 |
1.01824314 |
|
Moving |
0.516936946 |
0.159822846 |
3.234437128 |
0.001287 |
0.203047179 |
0.83082671 |
|
Course |
-2841.922435 |
323.8495507 |
-8.775440413 |
1.82E-17 |
-3477.958288 |
-2205.88658 |
|
Position<= o:p> |
3727.734432 |
612.7214704 |
6.083897191 |
2.11E-09 |
2524.35829 |
4931.11057 |
|
Public |
1017.487653 |
734.0216279 |
1.386182115 |
0.166214 |
-424.1202517 |
2459.09556 |
|
|
-2370.35675 |
853.0703272 |
-2.778618216 |
0.005633 |
-4045.774586 |
-694.938915 |
|
Country |
10433.07693 |
2433.056181 |
4.288054262 |
2.1E-05 |
5654.589633 |
15211.5642 |
|
Region |
667.4121287 |
478.8219254 |
1.393862923 |
0.163883 |
-272.987236 |
1607.81149 |
|
Degree |
710.1549718 |
898.2136745 |
0.790630328 |
0.429477 |
-1053.923747 |
2474.23369 |
|
Accreditation |
523.5253013 |
732.3294913 |
0.714876715 |
0.474968 |
-914.7592714 |
1961.80987 |
|
Year |
0.203605818 |
0.074203923 |
2.743868642 |
0.006256 |
0.057870382 |
0.34934126 |
APPENDIX
A
2002 Computer Professional Offer & Affirmative Action
Survey*
(Anonymous Form)
Please
fill out this anonymous form for each offer you
receive for academic year 2002-2003.
Privacy
Statement:
Please
be assured that there is NO WAY for anyone to discover your
identity. Please feel free to peruse the very simple and basic HTML to veri=
fy
that. Our site does NOT allow inspection of log files.
The Form
New Entry
Update or Correctio=
n
If this is a correction or update (for example, accepted off=
er,
higher offer), please provide the record# to change here-->
&nbs=
p;
(see the first col=
umn
of the table) and re-enter ALL information. This record will REPLACE the old
one.
About You
Your Last name
Your Day Time Phone Number
Your Cell Phone Number
Your Company Name
Your e-mail address
Your URL address
Number of years teaching experience you have, on an FTE
(full-time equivalent) basis (for example, two years of half-time teaching =
as a
doctoral student would count as one year).
Number of years non-teaching
experience you have, on an FTE (full-time equivalent) basis.
Your educational background (reasona=
bly
expected within a few months of your start date)
Country of national
origin of the professional *
Sex (Male or Femal=
e,
Other)
Race (White, Black,
Latino, Asian, Jewish, Indian, Other)
Marital Status
(Married, Single, Divorced, Other)
Religion (Christia=
n,
Muslim, Jewish, Hindu, Others)
Disabled (Yes, if
qualified according to the
Birth-Date
(10/30/1951)
Number Of Refereed
Journal Publication(s)
Number Of Non Refe=
reed
Journal Publication(s)
Number Of Book (Te=
xt
and Trade Books) Publication(s)
About Your Publications
Number of journal publications accep=
ted
(both in print and forthcoming) in journals in this list: (MISQ,
CACM, ISR, Management Science, JMIS, Decision Sciences, IEEE Transactions, =
HBR)
(these are journals appearing in the top 10 of over half of the scales show=
n in
the Saunders compilation on ISWorld, listed
above in the order in which they appear in the Saunders list.
Number of journal publications accepted=
in
other refereed journals in MIS not listed above (include chapters in refere=
ed
research books)
Number of textbooks published or forthc=
oming
Number of research-oriented (NOT text) =
books
published or forthcoming
Number of publications accepted in other
outlets (include refereed conferences and chapters in non-refereed books)
=
About
The Offer:
Base Salary (exclusive of summer
support) in US $. Those with 12-month contracts (usually due to administrat=
ive
duties) should not perform contrived calculations to derive summer support =
and
should report the entire salary here. Please do not enter text here.=
Guaranteed annual summer support in =
US $ ---for years
Annual discretionary research budget=
(including travel,
technology, and optional secretarial services, but not including required i=
tems
such as course software, basic telephone, and copying). If not specified
separately, estimate the maximum travel and technology that would reasonabl=
y be
supported without special requests (or appeals).
Moving expense reimbursements or sig=
ning
bonuses in US $
Annual teaching load in courses =
(semester-course
equivalents) defined as number of 3-credit semes=
ter
courses to be taught. A 3-credit course counts for roughly 3 hrs/week, so 15
weeks =3D nominally 45 hours per course inclusive of breaks. In practice, t=
aking
breaks will limit class time to about 50 minutes per class hour, or 2.5 hou=
rs
per week (37.5 teaching hours per semester). Thus, a 2+2 load at a
semester-based school would involve approximately 180 nominal hours (inclus=
ive
of breaks) or 150 teaching hours (exclusive of breaks), and would be entere=
d as
4. If you are not on the standard semester system, please divide
the number of nominal hours including breaks by 45, or divide the number of
teaching hours excluding breaks by 37.5, for a fairly clear answer to this
question. If the two differ, please use the latter. Feel free to use decima=
ls
in the answer, but please round to the nearest tenth.
Publication requirement for tenure at
this institution (leave blank if there is no tenure system).
Note: For the following items, it is
understandable that a school would indicate ranges rather than single numbe=
rs.
Please take a midpoint of the range you're thinking of, and assume a mix of=
33%
"A" publications and 66% "other" in a hypothetical pers=
on's
record. If ranges are provided, I will take the midpoint of each range
As best you can estimate, how many
top-tier publications (see definition above) will be required for tenure?
As best you can estimate, how many
publications overall (including the top-tier category) will be required for
tenure?
Country of the offering school
(If
Position for this
offer
Best describes this
particular school's campus
Highest
business/information science degree granted by the school that provides any
meaningful resources to its MIS faculty.
The type of
institution
Describes whether
unionization affects salary offer
Highest accreditat=
ion
of the school
Current status of =
this
offer
* Revised to measu=
re
disparities that discrimination may cause
APPENDIX B
Faculty Sa=
lary
Survey Analyzer & EEOC Affirmative Action Alert Variance Analysis=
(Us=
ing Sample
Faculty Data for
=
Tool by Ido Millet,
Data from the AIS/ISWorld Salary Survey Revised =
By
Rushinek & Rushinek
Data as of August 20, 2002
=
|
Filtering
Options: |
|
|
|
Country |
USA |
|
|
Accreditation |
National |
|
|
Teaching
/ Research |
Research
(<=3D4) |
|
|
Region |
US
SOUTH |
|
|
|
No |
|
|
Identity
Revealed? |
(All) |
|
|
Private
/ Public |
(All) |
|
|
Candidate
Education |
(All) |
|
|
Offer
Accepted? |
(All) |
|
|
US
Nationals |
Foreign
Nationals |
|
|
Results: |
|||||
|
XYZ University 2002-2003 |
|||||
|
Position |
Data |
US Nationals |
Foreign Nationals |
Foreign
- US Variance |
Foreign
- US Variance
% |
|
Tenured
|
Avg. Sal=
ary |
$110,000 |
$88,000 |
($22,000) |
-20% |
|
Associate |
%
w Research Budget |
0% |
|
|
|
|
|
Avg.
Research Budget |
$0 |
|
|
|
|
|
Avg.
Teaching Load |
4 |
6 |
2 |
+50% |
|
|
%
w Summer Support |
100% |
|
|
|
|
|
Avg.
Summer Support |
$18,000 |
|
|
|
|
|
Yrs
Summer Support |
2 |
|
|
|
|
|
Avg.
Moving Allowance |
$6,000 |
|
|
|
|
|
N |
1 |
1 |
|
|
|
|
|
|
|
|
|
|
Assistant |
Avg. Sal=
ary |
$106,000 |
$81,620 |
($24,380) |
-23% |
|
|
%
w Research Budget |
83% |
|
|
|
|
|
Avg.
Research Budget |
$5,500 |
|
|
|
|
|
Avg.
Teaching Load |
4 |
6 |
2 |
+50% |
|
|
%
w Summer Support |
100% |
|
|
|
|
|
Avg.
Summer Support |
$21,471 |
|
|
|
|
|
Yrs
Summer Support |
3 |
|
|
|
|
|
Avg.
Moving Allowance |
$6,954 |
|
|
|
|
|
N |
6 |
1 |
|
|
Switched
|
Avg. Sal=
ary |
$96,500 |
$76,230 |
($20,270) |
-21% |
|||
|
Assistant |
%
w Research Budget |
100% |
|
|
|
|||
|
|
Avg.
Research Budget |
$3,500 |
|
|
|
|||
|
|
Avg.
Teaching Load |
4 |
6 |
2 |
+50% |
|||
|
|
%
w Summer Support |
50% |
|
|
|
|||
|
|
Avg.
Summer Support |
$24,800 |
|
|
|
|||
|
|
Yrs
Summer Support |
2 |
|
|
|
|||
|
|
Avg.
Moving Allowance |
$3,000 |
|
|
|
|||
|
|
N |
2 |
1 |
|
|
|||
|
Summary
for All Positions Combined: |
|
|||||||
|
Data |
US Nationals |
Foreign Nationals |
Foreign
- US Variance |
Foreign
- US Variance
% |
|
|||
|
Avg. Sal=
ary |
$104,333 |
$81,950 |
($22,383) |
-21% |
|
|||
|
%
w Research Budget |
78% |
|
|
|
|
|||
|
Avg.
Research Budget |
$4,444 |
|
|
|
|
|||
|
Avg.
Teaching Load |
4 |
6 |
2 |
+50% |
|
|||
|
%
w Summer Support |
89% |
|
|
|
|
|||
|
Avg.
Summer Support |
$21,825 |
|
|
|
|
|||
|
Yrs
Summer Support |
3 |
|
|
|
|
|||
|
Avg.
Moving Allowance |
$5,969 |
|
|
|
|
|||
|
N |
9 |
3 |
|
|
|
|||
APPENDIX C
Total Dam=
age Revenues/Costs
Expected From a Discrimination Class Action Lawsuit
(Uses
the values for
|
Data |
Value |
Variable/Formula |
|
Expected
Class Action Revenues Per Foreign National Employee (Plaintiff) at |
||
|
Average Per Year $ Discrimination Damages |
$48,652 |
A (see
note at bottom) |
|
Average Number of Years of Discrimination |
10 |
B |
|
Total Recurring $ Discrimination Damages for All Years |
$486,520 |
C=3DA*B |
|
Average One Time Denied Moving Allowance |
$5,969 |
D |
|
Total $ Discrimination Damages for All Years Including Denied Movin=
g Allowance |
$492,489 |
E=3DC+D |
|
Simple Interest Rate on Damages for All Years of Discrimination |
5% |
F |
|
Simple Interest $ Amount on Damages for All Years of Discrimination=
|
$24,624 |
G=3DE*F |
|
Total $ Discrimination Damage Revenue for All Years |
$517,113 |
H=3DE+G |
|
Expected
Class Action Revenues for Multiple Plaintiffs and Institutions |
||
|
Average Number of Foreign Nationals Discriminated Against Per
Institution |
3 |
I |
|
Total Average $ Discrimination Damages at One Institution for All D=
iscriminated
Employees |
$1,551,339 |
J=3DH*I |
|
Number of Institutions Discriminating Against Employees Based on Na=
tional
Origin |
20 |
K |
|
Total Damage Revenue Expected From a Class Action Discrimination La=
wsuit |
$31,026,780 |
L=3DJ*K |
|
Expected Class Action Cash Outflows
Excluding Legal Costs (Lawyers) |
||
|
Web Server Setup, Configuration, & Programming Cost |
$500 |
M |
|
Hosting & Other Yearly Charges Per Web Server |
$500 |
N |
|
Number of Years of Web Server Usage |
10 |
O |
|
Total Web Server Hosting Cost for All Years of Usage |
$5,000 |
P=3DN*O |
|
Interest on Hosting Fees |
$250 |
Q |
|
Total Server Cost for All Years Including Interest |
$5,750 |
R=3DM+P+Q |
|
Expected Net Class Action Cash Inf=
lows
Excluding Legal Costs |
||
|
|
($31,021,030) |
=3DL-R |
Note: This value comes from the = 220;Summary for All Positions Combined” Table in Appendix B. It is the sum of (Fo= reign–US Avg. Salary) + (Avg. Research Budget) + (Avg. Summer Support).