H-1B Foreign Workers: Better Controls Needed to Help Employers and
Protect Workers (Letter Report, 09/07/2000, GAO/HEHS-00-157).

Pursuant to a congressional request, GAO provided information on the
Department of Labor's (DOL) H-1B program's implementation and
implications for the American workforce, focusing on: (1) the jobs that
H-1B workers are filling in the United States and the characteristics of
those workers; (2) the adequacy of the H-1B visa program's
implementation and enforcement; and (3) efforts underway to improve
information technology (IT) skills in the American workforce.

GAO noted that: (1) employers have used the H-1B visa program to fill
hundreds of thousands of positions in which certain skills, including
computer programming, engineering, education, and medicine, were needed
quickly; (2) according to Immigration and Naturalization (INS) data,
about 60 percent of the positions that new workers were approved to fill
in fiscal year (FY) 1999 were related to IT; (3) approved workers median
were scheduled to fill positions that offered initial annual salaries of
$45,000, workers had a median age of 28 years, and almost half were born
in India; (4) despite the H1-B program's success at helping employers
bring in highly skilled foreign workers, DOL's limited legal authority
to enforce the program's requirements and weaknesses in INS' program
administration leave the program vulnerable to abuse; (5) delays and
administrative problems have also lead to inefficient service for
employers using the program; (6) under the law, in certifying employers'
initial requests for H-1B workers, DOL is limited to ensuring that the
employer's application form has no obvious errors or omissions; (7) it
does not have the authority to verify whether information provided by
employers on labor conditions, such as wages to be paid, is correct; (8)
DOL has limited authority to ensure that employers are actually
complying with the law's requirements after the H-1B workers are
employed in the United States; (9) DOL generally cannot initiate
enforcement actions, even if it believes employers are violating the
law; (11) however, DOL agreed with GAO's matter to broaden DOL's
enforcement authority for the H-1B program; (10) INS is responsible for
ensuring that H-1B positions are in fact specialty occupations and that
workers granted entry are qualified for those positions; (11) however,
there is not sufficient assurance that INS reviews are adequate for
detecting program noncompliance or abuse; (12) INS decisions about the
priority of H-1B application processing related to other types of
petitions handled by INS have resulted in delays of several months to
process employers' requests for H-1B workers; (13) other system
weaknesses at INS have contributed to errors in counting the number of
visas approved under the H-1B visa program; and (14) to enhance U.S.
workers' ability to fill IT positions, DOL and the National Science
Foundation are working to improve the IT skills of the U.S. workforce.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-00-157
     TITLE:  H-1B Foreign Workers: Better Controls Needed to Help
	     Employers and Protect Workers
      DATE:  09/07/2000
   SUBJECT:  Labor force
	     Labor supply
	     Information technology
	     Labor law
	     Internal controls
	     Alien labor
	     Employment of foreign nationals
IDENTIFIER:  DOL H-1B Visa Program

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Testimony.                                               **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO/HEHS-00-157
Appendix I: Scope and Methodology

40

Appendix II: Labor's Technical Skill Grants and Examples of
NSF Grants Awarded

42

Appendix III: Comments From the Department of Labor

45

Appendix IV: Comments From the Department of State

52

Appendix V: Comments From the Immigration and
Naturalization Service

54

Appendix VI: Comments From the National Science Foundation

58

Appendix VII: GAO Contacts and Staff Acknowledgments

60

Table 1: H-1B Investigations, Violations, and Back Wages Due 22

Table 2: Labor's Technical Skill Grants Awarded, February 2000 42

Figure 1: New H-1B Nonimmigrants Approved, Fiscal Year
1992-Fiscal Year 1999 8

Figure 2: Summary of H-1B Visa Approval Process 9

Figure 3: Occupations for New H-1B Visas, Fiscal Year 1999 15

Figure 4: Age Distribution of Workers Approved for H-1B Visas,
Fiscal Year 1999 16

Figure 5: Countries of Birth for New H-1B Visa Approvals, Fiscal
Year 1999 17

ACWIA American Competitiveness and Workforce Improvement Act

ETA Employment and Training Administration

FLSA Fair Labor Standards Act

INS Immigration and Naturalization Service

IT information technology

LCA Labor Condition Application

NSF National Science Foundation

WHD Wage and Hour Division

Health, Education, and
Human Services

B-283330

September 7, 2000

The Honorable Patsy T. Mink
Ranking Minority Member, Subcommittee on
Criminal Justice, Drug Policy, and Human Resources
Committee on Government Reform
House of Representatives

Dear Ms. Mink:

A strong national economy depends, in part, on employers' ability to hire
workers with the necessary skills to perform needed tasks. Without these
workers, American businesses may be unable to sustain the economic
performance that has improved the quality of life for many Americans. The
information technology (IT) industry, in particular, is a driving force
behind current and future U.S. economic growth, and the Bureau of Labor
Statistics projects that the demand for workers with certain IT skills will
double over the next 10 years. Employers in the IT industry have expressed
concerns about not being able to fill their many vacancies. To help U.S.
employers in IT and other industries fill their needs for highly skilled
workers, the H-1B visa program allows employers to temporarily (for up to 6
years) fill needs in specialty occupations with foreign workers. Under the
law, H-1B workers must be employed in specialty occupations and have
suitable credentials for the job, and their employers must meet certain
labor conditions, including paying comparable wages.1 These requirements are
intended to ensure that American workers are not adversely affected.

The number of foreign workers legally authorized to enter the United States
annually through the H-1B program has increased substantially--from 65,000
in 1992 to 115,000 in 1999 and 2000. Some believe the limit should be raised
even further to address workforce needs, such as for IT workers. However,
others question whether enough is being done to increase the skills of
American workers so they can fill these vacancies. Also, some employers are
dissatisfied with how the Department of Labor and the Immigration and
Naturalization Service (INS) administer the H-1B program.2 Finally, reports
of program misuse--such as employers paying workers less than comparable
wages or employees using false credentials--have led to questions about
whether the program adequately serves employers or protects workers.

Because of these concerns, you asked us to provide information on the H-1B
program's implementation and implications for the American
workforce--specifically (1) the jobs that H-1B workers are filling in the
United States and the characteristics of those workers, (2) the adequacy of
the H-1B visa program's implementation and enforcement, and (3) efforts
underway to improve IT skills in the American workforce. To answer these
questions, we obtained and independently analyzed newly collected INS data
on approved H-1B workers' characteristics; observed and evaluated
application processing at several INS and Labor offices around the country;
obtained and evaluated available processing data; and discussed the
program's history, current operations, and limitations with INS, Labor,
State, and National Science Foundation (NSF) officials, employer and
employee associations representing numerous IT employees and employers, the
IT industry and users of the H-1B program, and 13 individual employers,
primarily in the IT industry. We conducted our work in accordance with
generally accepted government auditing standards between August 1999 and
July 2000. (See app. I for a full discussion of our scope and methodology.)

Employers have used the H-1B visa program to fill hundreds of thousands of
positions in which certain skills, including computer programming,
engineering, education, and medicine, were needed quickly. According to INS
data, about 60 percent of the positions that new workers were approved to
fill in fiscal year 1999 were related to IT. Workers approved for H-1B visas
were scheduled to fill positions that offered initial median annual salaries
of $45,000. The workers had a median age of about 28 years at the time of
approval, and almost half were born in India. About 40 percent of them were
already in the United States on another type of visa. Those workers approved
for H-1B visas in IT-related occupations differed somewhat from other H-1B
workers in that they were less likely to have an advanced degree, were
younger, more likely to be from India, and less likely to be in the United
States on another type of visa when approved for the H-1B program.

Despite the H-1B program's success at helping employers bring in highly
skilled foreign workers, Labor's limited legal authority to enforce the
program's requirements and weaknesses in INS' program administration leave
the program vulnerable to abuse. Delays and administrative problems have
also led to inefficient service for employers using the program. Under the
law, in certifying employers' initial requests for H-1B workers, Labor is
limited to ensuring that the employer's application form has no obvious
errors or omissions. It does not have the authority to verify whether
information provided by employers on labor conditions, such as wages to be
paid, is correct. Moreover, some of this same information is reviewed again
by INS during its assessment of employer requests for workers. Further,
Labor has limited authority to ensure that employers are actually complying
with the law's requirements after the H-1B workers are employed in the
United States. Unlike under other labor laws it enforces, Labor generally
cannot initiate enforcement actions (such as conducting investigations and
subpoenaing employer records), even if it believes employers are violating
the law. We have included two matters for congressional consideration to
address Labor's limited authority under the law. Labor disagreed with our
matter concerning the transfer of LCA review to INS, arguing that
consideration should be given to making Labor's review more substantive.
Unless the Congress decided to authorize a more substantive review, transfer
to INS would be more efficient. Labor agreed with our second matter to
broaden Labor's enforcement authority for the H-1B program.

INS is responsible for ensuring that H-1B positions are in fact specialty
occupations and that workers granted entry are qualified for those
positions. Until recently, INS had no national systematic approach for
adjudicators to follow to ensure the consistent review of employer
petitions. Further, INS staff conducting these reviews continue to lack easy
access to specific, case-related information that would help them assess the
merit of employers' requests, which can also lead to inconsistent or
incorrect approvals of requests. Because existing supervisory review and
performance appraisal systems for INS staff reviewing petitions are based
largely on the number of requests processed, rather than the quality of the
review, staff can be rewarded for timely handling of petitions rather than
for careful scrutiny of petitions' merits. As a result, there is not
sufficient assurance that INS reviews are adequate for detecting program
noncompliance or abuse. In addition, INS decisions about the priority of
H-1B application processing related to other types of petitions handled by
INS have resulted in delays of several months to process employers' requests
for H-1B workers. Other system weaknesses at INS have contributed to errors
in counting the number of visas approved under the H-1B visa program. We
make three recommendations designed to enhance the consistency and
correctness of INS' H-1B decisions. INS generally did not agree with our
recommendations, believing that current program procedures are sufficient to
detect noncompliance and abuse. We continue to believe that actions beyond
those taken by INS are warranted.

To enhance U.S. workers' ability to fill IT positions, Labor and NSF are
working to improve the IT skills of the U.S. workforce. The IT employers we
contacted told us that they are also trying to improve U.S. workers' IT
skills and identified a variety of short-term methods, such as retraining
new or existing employees, to provide U.S. workers with the needed skills.
On a longer-term basis, IT employers reported using a variety of programs to
encourage U.S. students to pursue IT careers, such as providing computer
training and mentoring for students in elementary and secondary schools.
Efforts by Labor and NSF, funded through fees paid by employers wishing to
use the H-1B program, include training grants and scholarships. These
efforts should help increase the number of U.S. workers with IT skills.
However, their ultimate effect is unknown because the programs are new,
their focus is longer-term, and in some cases, there is a lack of data about
what IT skills are needed.

The H-1 nonimmigrant category was created under the Immigration and
Nationality Act of 1952 (P.L. 82-414) to assist U.S. employers needing
workers temporarily. The Immigration Act of 1990 (P.L. 101-649) amended the
law by, among other things, creating the H-1B category for nonimmigrants who
are sought to work in specialty occupations, and fashion models "of
distinguished merit and ability."3 Unlike most temporary worker visa
categories, H-1B workers can intend both to work temporarily and to
immigrate permanently at some future time.

No limit existed on the number of specialty occupation visas that could be
granted until 1990. Through the Immigration Act of 1990, the Congress set a
yearly limit of 65,000 for H-1B visas alone, which took effect in fiscal
year 1992. In an effort to help employers access skilled foreign workers and
compete internationally, in 1998 the Congress enacted the American
Competitiveness and Workforce Improvement Act (P.L. 105-277, Title IV)
(ACWIA), which increased the limit to 115,000 for fiscal years 1999 and
2000, and 107,500 for fiscal year 2001. The limit is scheduled to revert to
65,000 in fiscal year 2002. The number of visas approved did not reach the
annual limit until fiscal year 1997 (see fig. 1), and exceeded the limit in
fiscal year 1999 by more than 20,000.4 In March 2000, INS stopped accepting
new petitions for fiscal year 2000, believing it had received enough to
reach the limit. Legislation has been introduced in the Congress to further
increase the number of H-1B workers authorized to enter the United States or
to eliminate the limitation entirely.5

Source: INS.

An employer who wishes to hire an H-1B worker must follow several steps,
beginning with the submission of a labor condition application (LCA) to
Labor (see fig. 2).6

Note: At each step of the process, the application/petition could be denied;
employers have the ability to resubmit their forms or appeal such decisions.

On the LCA, the employer must identify the number of workers requested and
the occupation and location(s) in which they will work, and show the wages
that they will receive. The employer must attest that

ï¿½ the employment of H-1B workers will not adversely affect the working
conditions of other workers similarly employed in the area;

ï¿½ the H-1B workers will be paid wages that are no less than the higher of
the actual wage level paid by the employer to all others with similar
experience and qualifications for the specific employment or the prevailing
wage level for the occupational classification in the area of intended
employment;

ï¿½ no strike, lockout, or work stoppage in the applicable occupational
classification was underway at the time the application was prepared;

ï¿½ a copy of the application will be given to the H-1B worker; and

ï¿½ a copy of the application will be given to the employees' bargaining agent
or, if there is no bargaining agent, will be posted for a 10-day period to
inform potentially affected workers of the application for an H-1B worker.
This helps to inform workers of their right to file complaints if they do
not believe the employer is complying with the law.

The employer must maintain public files to support and document the
information attested to on the application.

In an effort to protect U.S. workers, ACWIA requires employers who are H-1B
dependent (generally those with a workforce consisting of at least 15
percent H-1B workers) to provide additional information and comply with
additional requirements regarding recruitment and layoff procedures. These
additional requirements, which will not apply to most applications after
September 30, 2001, have not gone into effect because, according to a Labor
official, implementing regulations are still awaiting organizational
approval.

After Labor approves the LCA, the employer then files a petition (referred
to as the I-129), along with the approved LCA with INS.7 The petition is
required to contain the necessary information to show that a bona fide job
exists and that the prospective H-1B worker has the requisite education and
work experience for the position. The employer must file a petition for each
H-1B worker. Information provided includes the type of business, the
employer's income, the number of employees, and the prospective worker's
educational background or work experience. INS staff review the petition and
LCA, ensuring that the petition information indicates that the labor
conditions on the LCA will be met, that a bona fide job exists for the
worker, and that the worker meets the qualifications for the designated
specialty occupation. With each petition, the employer submits a filing fee
of $110; ACWIA authorized an additional $500 fee to be used for skill
grants, scholarships, and other purposes (as discussed later), although some
types of employers are exempt from this.

If the petition is approved, INS notifies the Department of State to issue a
visa for the worker. State's consular officers abroad review these petitions
to assess the potential worker's visa eligibility. The consular offices
generally interview the potential worker to decide whether to issue a visa,
and, if appropriate, do so. If the worker is already in the United States in
another visa status (such as a student), the worker applies to INS to change
the visa status. Petitions are approved for up to a 3-year period; employers
can apply for extensions, but H-1B workers are limited to a 6-year maximum
stay.

After the H-1B worker is employed, Labor and INS also have responsibility
for ensuring that the employer is complying with program requirements and
that both H-1B workers and their American counterparts are receiving the
protections guaranteed under the law. Labor's Wage and Hour Division (WHD)
is the Labor entity responsible for enforcing a number of labor laws
governing wages and working conditions, including the Fair Labor Standards
Act (FLSA), which governs minimum wage, child labor, and overtime pay. To
assess compliance with FLSA as well as other laws, WHD conducts on-site
investigations based on a variety of criteria, obtains records from
employers (and may subpoena such records), and may cite and fine an employer
for noncompliance. It also surveys particular industries and employers
within them to obtain a baseline understanding of compliance, then targets
either employers or specific industries for further investigations to
address the most serious abuses. For the H-1B program, WHD is responsible
for ensuring that H-1B workers are actually working in the occupation listed
on the LCA and receiving the promised wages. INS is responsible for
detecting visa fraud across all visa categories and conducting
investigations based on a number of criteria. INS reviews whether worker
qualifications are appropriately represented, the employer has provided a
job for the worker, the employer is using the worker in the specialty
occupation, and whether the wage promised in the petition is being paid.
Enforcement authority includes citing and fining employers, revoking the
petition approval, seeking prosecution if criminal statutes are violated,
and possibly seeking the removal of aliens if violations are found.

Although employers are not required to prove a shortage of U.S. workers
exists in order to recruit H-1B workers, there has been a longstanding
debate about whether, in fact, sufficient numbers of U.S. workers with the
requisite skills to fill current IT vacancies are available. These debates
have led to numerous studies, but definitional and methodological problems
in these studies do not permit a conclusion as to the extent of any IT skill
shortage. Studies have estimated IT vacancies in the United States from
190,000 to more than 700,000, but some studies define IT workers very
broadly whereas others focus on specific IT occupational clusters. Moreover,
the studies provide little information about these vacancies, such as how
long positions were vacant, whether sufficient wages to attract workers were
offered, or whether companies considered jobs filled by contractors as
vacancies. We, the Department of Commerce, and the Computer Research
Association (under an NSF grant) have reported that more information is
needed to characterize the IT labor market and determine the extent of any
shortage.8 In October 2000, NSF plans to report its assessment of labor
market needs for IT workers; however, according to an NSF official, the
report will provide anecdotal information rather than empirical evidence.

This lack of data has caused some to be concerned about employers'
increasing use of the H-1B program. Associations representing U.S. workers
believe that IT employers have exaggerated the need for foreign workers,
which could lead to a surplus of workers in certain occupations and depress
wages for all workers in the long term. Officials representing these
associations cite instances in which IT employers have hired younger,
temporary foreign workers with narrowly focused skills rather than older
U.S. workers who could be retrained. They also note that if, indeed,
employers cannot find skilled U.S. workers, they could hire foreign workers
through permanent employment-based immigration programs. However, this
process requires documentation indicating a particular need for each worker,
and can take several years. Moreover, there are limits on the number of such
visas that can be granted each year.

The debate has also led to questions about whether enough has been done to
identify and train U.S. workers to fill IT vacancies. As a result, various
federal efforts have been implemented to improve the IT skill set of
workers.9 As noted above, the most recent legislation required most
employers wishing to use the H-1B program to pay a $500 fee that would
largely be used to fund two efforts to increase the skill set of American
workers. The first effort, administered by Labor, provides training through
skill grants. The second effort, administered by NSF, funds scholarships
through universities to encourage low-income students to enter fields of
computer science, engineering, and math. Of the total fees collected from
employers wishing to use the H-1B program, Labor will receive 56.3 percent

for the technical skill training grants and NSF will receive 28.2 percent
for scholarships. The legislation provides other uses for the remaining
amounts.10

In the past, few accurate data have been made available about the jobs
filled by and characteristics of workers approved for H-1B visas. However,
in 1999, for the first time, INS compiled information on key characteristics
such as occupations, wages, degrees earned, and countries of birth from a
statistical sample of new H-1B visa approvals.11 INS has now begun to
regularly collect and analyze specific information on new H-1B workers, as
well as renewals, as required under ACWIA. It still lacks important
information on these workers, however, such as an accurate, unduplicated
count of approved H-1B workers who actually enter the United States on an
H-1B visa.

In 1999, workers approved for H-1B visas were to fill a wide array of
specialty occupations. As shown in fig. 3, 59 percent of the visas were for
workers slated to work in IT-related occupations, including the large

category of systems analysis and programming.12 Another 5 percent were in
electrical or electronics engineering occupations, fields that may also be
related to computer development.13 The remaining visas were for occupations
in fields as diverse as college and university education, accounting and
auditing, biological sciences, economics, mechanical engineering, medicine,
and commercial art. INS has few historical data for comparison, but in
fiscal year 1992, about 6,000 H-1B visas were approved for IT-related
occupations--many fewer than the approximately 79,000 estimated to have been
approved for IT-related occupations in fiscal year 1999.

Note: Totals do not equal 100 percent due to rounding.

Source: INS.

The majority--57 percent--of workers approved for new H-1B visas in fiscal
year 1999 had earned a bachelor's degree as their highest degree, while 41
percent had attained an advanced degree. The positions they were scheduled
to fill offered a median initial annual salary of $45,000. About 35 percent
of those workers approved for IT-related occupations had advanced degrees,
as compared with 50 percent for those approved for the remaining
occupations.

Workers approved for new H-1B visas in 1999 had a median age of 28.3 years
(see fig. 4); over 80 percent were younger than 35 years old. Those workers
approved for IT-related occupations were younger than their non-IT
counterparts--a median age of 27.4 years, compared with 30.2 years.

1999

Source: INS.

Almost half of the workers approved for new H-1B visas in fiscal year 1999
were born in India, with the second highest number of workers born in China
(see fig. 5). In total, at least 119 countries were represented. Almost
three-quarters of the workers approved for IT-related occupations were born
in India, compared to 14 percent for those workers approved for
non-IT-related occupations.

Source: INS.

About 42 percent of those workers approved for H-1B visas in 1999 were
already in the United States when their visas were approved. The majority of
those here were on student visas, while others were spouses or children of
foreign students, visitors, or other types of nonimmigrants. About
one-quarter of the workers scheduled to fill IT-related occupations were
already in the United States, which indicates that H-1B workers for
IT-related occupations are more likely than those in non-IT related
occupations to be recruited from outside the country.

Unavailable

As required under ACWIA, INS has been collecting specific data on H-1B visa
petition approvals for fiscal year 2000. Although INS had previously
obtained some of these data on the petition or supporting documentation
(such as workers' country of origin, educational levels, and wages to be
paid), it did not routinely analyze all of this information. It is now
collecting and analyzing this information in addition to other information,
such as the worker's major or primary field of study and the employer's
industry code, in an effort to shed light on characteristics of H-1B
visa-holders and the jobs they fill in the United States.

Preliminary data from new visa approvals as of February 29, 2000, mirror the
1999 data. For example, nearly 50 percent of the petitions counting against
the annual limit--those for new workers--were for IT-related occupations (as
compared with 59 percent from 1999). INS also is finding that about 60
percent of approvals not counting against the limit were for IT-related
occupations, indicating that workers in computer-related occupations are
more likely to extend their stay in the United States, move between
companies, or work concurrently for another employer than other H-1B
workers.

However, INS is still unable to determine the number of H-1B workers
approved in any given year who actually come to the United States. This
inability to accurately identify the number of workers who actually enter
the country on a particular visa or who stay in the country after their
initial term has ended, obtain a visa extension, or obtain permanent
residency extends beyond the H-1B program. For example, in 1998, about
13,000 H-1B nonimmigrants became legal permanent residents, but there are no
data on how many returned to their home country, or stayed here illegally
after their H-1B visas expired.14 This is because INS has two systems that
do not interact with each other--one that tracks the number of visa
petitions it approves, and another that tracks nonimmigrants entering the
country on visas. Nonimmigrants may enter and leave the country several
times during the period of their visa but these entrances are not matched
with the year of approval, nor does INS attempt to eliminate repeat
entrances in its nonimmigrant system statistics. The Congress recognized the
need for better information on nonimmigrants entering and leaving the
country and in 1996, through the Illegal Immigration Reform and Immigrant
Responsibility Act, required INS to develop an automated entry/exit control
system that would provide better records on every individual arriving in or
departing the United States under a visa. It was anticipated that, by having
such records, INS would be able to link individual nonimmigrants with their
particular visa category. In June 2000, the Congress extended the full
implementation date for this system to December 31, 2005.

Lead to Inefficient Customer Service

Labor has limited legal authority for questioning an LCA and initiating
enforcement actions, such as investigations, to address potential
noncompliance. Moreover, INS lacks the necessary program controls to ensure
that each petition is correctly reviewed in a timely manner. As a result,
the program is vulnerable to abuse--both by employers who do not have bona
fide jobs to fill or do not meet required labor conditions, and by potential
workers who present false credentials. In addition, employers who meet H-1B
requirements may not be able to obtain the H-1B workers they need in a
timely manner. Finally, systems weaknesses at INS also lead to INS'
difficulty in accurately tracking the number of visas counted against the
annual limit.

Under the law, Labor's Employment and Training Administration (ETA) has 7
days to decide whether to certify the conditions that an employer attests to
on the LCA.15 Because the law permits Labor to review the LCA "only for
completeness and obvious inaccuracies," ETA cannot evaluate other types of
information, regardless of how questionable it may appear. As a result, ETA
approves most LCAs--93 percent in fiscal year 1999 (and, according to a
Labor official, many of the LCAs initially rejected were likely resubmitted
with changes and then approved). While this part of the H-1B application
process was established to require the employer to attest to the existence
of various labor conditions, the Congress wanted it to be as quick and
efficient as possible by limiting the depth and time frame of Labor's
review.

However, this statutory limitation means that Labor's certification can do
little to ensure that employers are meeting required labor conditions. For
example, an employer must certify that no strike, lockout, or work stoppage
was under way at the time the LCA was prepared; however, according to ETA
officials, even if they find out through other means that a strike is under
way, they must approve the LCA. As another example, even though employers
may be required to pay H-1B workers a prevailing wage, ETA officials said
employers can use almost any source to determine a prevailing wage and ETA
does not have the authority to verify the authenticity of the information
unless officials can demonstrate that the source is obviously inaccurate on
its face. According to ETA officials, even if they know a prevailing wage is
incorrect, they must approve the LCA.

Once Labor has certified the LCA, the employer must provide it, along with
the petition, to INS. The petition includes some of the same information as
the LCA, and, as a part of its review, INS reviews information on both
documents (such as the wages to be paid). In that respect, the filing of the
LCA with Labor represents an additional step for employers that adds
additional effort and at least a week to H-1B processing time.

Labor's WHD has limited ability to ensure that H-1B employers comply with
their legal obligations and H-1B workers and their U.S. counterparts are
protected under the law. When authorized to investigate, WHD is responsible,
under the H-1B law, for ensuring that workers are receiving the wages
promised on the LCA and are working at the occupation and location
specified; it can only initiate H-1B-related investigations as a result of
one of three factors:

ï¿½ A complaint is received from an aggrieved person or organization, such as
the H-1B worker, an American worker, or the employee bargaining
representative. Information that surfaces from ETA's or INS' review of an
employer's information on an LCA or INS petition is prohibited under the law
from being used as a basis for compliance investigations. WHD receives few
complaints--135 in fiscal year 1999--yet about 137,000 H-1B workers were
approved that year. A Labor official told us that these workers are
reluctant to complain about their working conditions, as they are dependent
on employers to enable them to stay in the United States or sponsor their
permanent residency.

ï¿½ WHD obtains information about a particular employer who, within the last 5
years, has been found to have committed a willful failure to meet a
condition specified in the LCA or willfully misrepresented a material fact
in the LCA. According to WHD officials, these criteria are very difficult to
meet in order to sustain legal challenges by employers and, to date, WHD has
only established one employer as a "willful" violator.

ï¿½ WHD receives specific credible information from a reliable source (other
than a complainant) that the employer has failed to meet certain specified
LCA conditions, has engaged in a pattern or practice of failures to meet
such conditions, or has committed a substantial failure to meet such
conditions that affects multiple employees.16 The Secretary of Labor must
personally certify that these conditions exist. WHD has yet to receive any
specific credible information that could justify an investigation under the
Secretary's special authority.

These limitations contrast with WHD's enforcement authority under other
worker protection laws, in particular, FLSA. As mentioned earlier, under
FLSA, WHD can initiate inspections on a variety of criteria to determine
potential noncompliance. It can survey industries to obtain a baseline
measure of the extent of compliance. It can then use that information to
target either employers or specific industries for further investigations,
thereby directing its investigation resources to eliminate the greatest
program abuses. Further, WHD during investigations can subpoena the
necessary records from employers, such as payroll documents, to determine
whether employers are paying the appropriate wages. Under the H-1B program,
however, WHD has no authority to perform these activities.

In many ways H-1B workers are different from workers covered under FLSA,
which affects WHD's ability to use other laws to ensure H-1B workers receive
their legal protections. First, FLSA's protections are focused on workers
who are lower paid, which H-1B workers generally are not. Second, WHD
investigators are more likely to investigate companies that have many
lower-paid workers, which are not the typical companies that use H-1B visa
workers. Finally, during investigations, WHD investigators would not likely
know which employers are H-1B workers because that kind of documentation is
not typically available during investigations. However, in a very important
way, H-1B workers are very much like workers covered under FLSA; according
to a Labor official, H-1B workers may be vulnerable to abuse since their
dependency upon their employers may lead to reluctance to complain, not
unlike those workers protected under FLSA. As a result, according to WHD
officials, the original assumption that enforcement for H-1B workers could
be complaint-driven has not held true.

Although its authority to investigate is limited, there is evidence to
believe that program noncompliance under the H-1B program exists. For
example, even though there has not been a large number of complaints, WHD is
significantly more likely to find violations in H-1B complaint cases than in
complaint cases under other laws, according to WHD officials. As shown in
table 1, over the last 4 1/2 years, 83 percent of the closed H-1B
investigations found violations−compared to about 40 to 60 percent
under other labor laws, according to Labor officials, and the amount of back
wages owed to H-1B workers has been substantial--over $2 million, or about
$3,800 per employee found to have back wages due.

       Investigations completed        Back wages due
 Fiscal                                                          Number of
 year  Total Violations Violations     Investigations Amount due employees
             (number)   (percentage)                  (dollars)
                                                                 involved
 1996  24    22         92             20             $335,454   36
 1997  16    14         88             14             149,458    33
 1998  24    20         83             17             365,840    90
 1999  32    24         75             19             361,534    206
 2000a 35    29         83             29             1,168,154  255
 Total 131   109        83             99             $2,380,440 620

Source: WHD, Department of Labor.

aThrough April 30, 2000.

Finally, according to WHD officials, there are increasing instances of
program abuse in which workers are brought into the United States to work,
but are not employed and receive no pay until jobs are available (often
called "benching"). Other violations have included employers withholding
wages from employees who have voluntarily left for employment elsewhere.
WHD's investigative findings are corroborated by a 1996 Labor Inspector
General report that found 75 percent of the aliens were working for
employers who did not adequately document the proper wage on the LCA and,
when the actual wage could be determined, 19

percent of the H-1B workers were paid less than the wage specified on the
LCA.17

INS staff, called adjudicators, review employers' petitions and decide
whether to approve the nonimmigrants' H-1B visa classifications. They review
the petitions and supporting documentation to determine whether bona fide
jobs exist for the H-1B workers--that is, jobs that meet the requirements of
a specialty occupation. They also are supposed to determine whether the
petition indicates that the qualifications of the prospective H-1B workers
meet the statutory requirements--for example, that they have a bachelor's or
higher degree (or its equivalent) in the specific specialty. They also
compare the information on the petition with that provided on the
Labor-certified LCA. During our review, we found that adjudicators did not
have a systematic approach for reviewing petitions. INS recently implemented
national standard operating procedures with criteria for adjudicators to
follow. Adjudicators continue to lack easy access to case-related
information that could help them make decisions about the merit of the
petitions. We found that the supervisory review and performance appraisal
processes give adjudicators incentives to approve petitions. These
procedures leave the program susceptible to abuse. We also found that
decisions on work priority at the various service centers have led to delays
of several months in reviewing employers' H-1B visa petitions.

Until Recently, INS Lacked a National Systematic Approach for Petition
Review

Before August 2000, INS had no national systematic approach for how
adjudicators at all four INS service centers were to determine whether an
employer's request for workers should be approved. Service center officials
and adjudicators said that although they initially received national-based
training, and there were center-based standard operating procedures, this
training or center-based guidance did not provide them the kind of practical
information they needed to assess petitions. In interviews with adjudicators
and observations of them performing reviews, we found major differences
among and within the different centers in how adjudicators decide which
petitions to approve.

For example, at one service center, employers were required to provide a
report from an independent agency that specializes in evaluating the
educational or work credentials of foreign workers; other service centers we
visited did not have such a requirement. Moreover, even within the service
center that required these reports, one adjudicator said that she would not
accept reports from a particular agency, while another adjudicator said that
he would accept this agency's reports. Also, an adjudicator at one center
said that, because of her experience in the high-technology field, she
questions whether certain IT occupations should be considered specialty
occupations under the H-1B program, and she would apply this belief to her
review of petitions. Adjudicators at other centers accepted these
occupations more readily.

We also found a wide range of views among the service centers as to whether
and how adjudicators may request additional information from an employer to
assess the merit of the petition.18 Adjudicators at each service center said
they were not sure whether the law allowed them to request additional
information, especially concerning whether a bona fide job exists and, as a
result, were reluctant to do so. However, an adjudicator at one service
center said that an adjudicator can request any information he or she
believes is necessary to assess whether such jobs exist and has instructed
adjudicators on how to obtain this information from employers.

INS headquarters officials said these differences may result from
adjudicators' discretion, or from differences in petitions that may not be
initially apparent. However, it acknowledged that gains could be made in the
efficiency and consistency of reviews and, in August 2000, implemented a
75-page set of national standard operating procedures that established a
systematic approach for adjudicators to use when reviewing H-1B petitions.
INS officials said that, in conjunction with these new procedures, it
conducted national training at all centers on the H-1B program. The standard
operating procedures lay out the basic steps that adjudicators should follow
when reviewing petitions, and explain the types of documentation that should
accompany the petition. The procedures require adjudicators to document why
they denied a petition or why they believed additional information was
necessary. The implementation of these procedures should help all
adjudicators understand what steps they should follow to assess the merit of
petitions. However, the procedures do not clearly detail how adjudicators
will address several of the issues we identified, such as how an adjudicator
will decide on the sufficiency or accuracy of the documentation provided, or
the criteria for, or situations under which, an adjudicator should request
additional documentation from an employer or deny a petition. Moreover, the
procedures do not require any documentation of the process used by
adjudicators to approve a petition, which account for the vast majority of
petitions reviewed. Without some explanation of those cases that are
approved, it is difficult for supervisory reviewers to determine whether
adjudicators actually took the appropriate steps when approving petitions.
According to INS, the procedures will continue to evolve to address
adjudicators' needs.

INS has also begun to compile relevant decisions in H-1B court cases that
will help adjudicators make decisions consistent with past, binding cases.
While these cases are helpful, service center officials noted that they
would like better training on how to use judicial precedents and how to word
requests for additional information that would improve adjudicators'
effectiveness.

Case-Specific Information That Could Help Adjudicators Is Not Easily
Accessible

Adjudicators commented that they do not have easy access to case-specific
information developed by other INS officials that would help them better
determine whether a petition should be approved. As INS' information systems
now operate, information that supports petition denials, such as evidence of
a fraudulent employer or falsified worker credentials, is either not
available to adjudicators, or not available in a manner that is easily
accessible for adjudicators, given the timeframes in which they must review
petitions. Information that a petition has been denied is initially only
available to adjudicators within the same service center. After a month, it
is uploaded to a central system and is available to adjudicators in all
centers, but can be accessed only through a complicated, time-consuming
process. Even if an adjudicator accesses the information, the reason for the
denial is not recorded in the automated file, so the adjudicator cannot
readily use the information to assess a petition from the same employer or
for the same potential employee. As a result, a petition previously
submitted and denied can be approved by another adjudicator, even if the
denying adjudicator determined that the employer does not meet H-1B
requirements. According to an INS official, in addition to information
developed by adjudicators, results of INS investigations of employers
already approved for H-1B workers are not readily available to adjudicators
on the information system. Adjudicators noted that they are under pressure
to adjudicate cases quickly and, unless the information is accessible on an
automated basis, they do not have time to review it. Such information is
sometimes available from service center staff who focus on fraud
investigations, but adjudicators and fraud staff at the centers explained
that because of the fraud staff's many responsibilities and adjudicators'
time pressures, adjudicators would not routinely use investigations staff to
look into potentially fraudulent petitions.

Officials at INS headquarters acknowledged that adjudicators need timely,
accurate, and accessible information in order to properly assess the merit
of petitions. These officials said that INS is in the process of stabilizing
and upgrading the petition-tracking computer system to correct some other
problems and believes that the upgrade could enable them to make better
information available on-line for adjudicators.

Supervisory Review and Performance Appraisal Processes Encourage Approvals

The process for assessing adjudicators' performance can give adjudicators an
incentive to approve petitions rather than scrutinize them carefully for
their merit. Currently, supervisors are required to routinely review only
denials and any requests that adjudicators might make for additional
information. They generally do not review approvals, which represent about
91 percent of petitions reviewed for all nonimmigrant employment visas. As a
result, in order to reduce the amount of supervisory review, an adjudicator
may approve petitions rather than deny them. Further, according to service
center officials, INS' current performance appraisal system for adjudicators
is based on the number of petitions reviewed, not the quality of the review.
They said that staff who process the greatest number of petitions are
generally rewarded over those who tend to assess petitions more critically
and, therefore, review fewer petitions in a given time period. They added
that while adjudicators understand that their responsibility is to carefully
review petitions, the performance appraisal system provides disincentives to
deny a petition or request additional information because of the additional
time it will take to reach a decision. INS headquarters officials explained
that the absence of a quality measure in performance appraisals reflects the
difficulty of implementing a reasonable measure, not an encouragement of
production over quality. INS officials said that, at various times, INS has
experimented with including a quality measure in the adjudicator performance
work plan but the various approaches have had significant drawbacks.

Implications of Inadequate INS Adjudication

The combination of the lack of guidance in particular areas, difficulties in
accessing case-related information, and the performance assessment
procedures have left the program vulnerable to program noncompliance and
abuse by potentially allowing H-1B petitions that do not meet requirements
to be approved. There is evidence that some employers and workers have tried
to abuse the program. INS investigators following up after petitions have
been approved have found a number of instances of program fraud in the
program. For example, INS has found workers brought to the United States
under the program who worked in occupations that did not qualify as H-1B
occupations; it has also identified employers who have created shell
corporations and created false credentials and documents for aliens who were
not eligible for H-1B employment. In 1998 and 1999, INS referred petitions
to the State consular post in Chennai, India, if they had certain fraud
indicators, such as a degree from a university often used in forged degrees.
State found that of the 3,247 petitions referred through March 31, 1999,
close to 45 percent of claims made on these petitions were of questionable
validity and 21 percent of the work experience claims made to INS were
fraudulent.19

In addition to eliminating potential program noncompliance or fraud, INS'
review of petitions needs to be effective for several other reasons. First,
INS typically does not verify whether the workers it approves actually work
in the jobs for which it approved the petitions and, according to INS
officials, detection of visa fraud after petitions have been approved is not
an investigative priority because limited special agent resources are, of
necessity, primarily devoted to criminal activities. Second, the State
Department's consular offices are generally required to interview each
applicant, but can waive this requirement when the consular office is
satisfied (based on a review of the application) that the applicant
qualifies for a visa. It relies on INS to ensure that petition information
related to U.S. employers is correct. One State Department foreign officer
said he assumes that Labor and INS, respectively, have satisfied themselves
on these issues before approving the LCAs and petitions. Third, we believe
that because the number of visas that may be issued for H-1B workers each
year is limited, there should be procedures in place to ensure that these
visas are granted judiciously and correctly so that those eligible for the
program have access to the limited visas.

The procedures can also lead to inconsistent reviews across and within
service centers that frustrate employers and prospective workers alike. One
adjudicator told us that she had denied a petition, and when the

employer resubmitted it without modifications, another adjudicator at the
same center approved it.

INS' Prioritization Decisions Lead to Delays for H-1B Processing

Before February 2000, INS had an established time frame of 30 days within
which adjudicators should start their review of H-1B petitions.20 At that
time INS increased the time frame to initiate H-1B petition review to 60
days. According to INS officials, the revision in processing times was an
effort to balance priorities and workloads at the service centers. At that
time, processing times on other types of petitions, which make up the
majority of petitions INS reviews, had grown to a year or more.21

As a result of this decision, at one service center we found petitions
waiting in a file room for 2 months or more before being distributed to
adjudicators for review. Moreover, we found in April 2000 that the four
centers were exceeding the 60-day time frame, taking anywhere from 45 to 70
days to start reviewing petitions. According to INS headquarters officials,
however, as of August 2000, the service centers were generally taking 60
days or less to process petitions.

Although INS' decision was made to address greater priorities, it
nonetheless has led to delays for employers using the H-1B program, and,
according to employers, has affected their ability to staff projects when
workers are needed. Employers said that, although the time it takes to hire
an H-1B worker varies, the LCA and INS petition process can exceed 4 months.

In 1999, INS discovered that it had approved more than the allowable number
of H-1B visas for fiscal year 1999. Recognizing the need to determine the
extent of the overage, INS engaged KPMG Consulting, LLC, to estimate the
number of H-1B petitions approved by INS that applied to the 1999 limit,
document the current H-1B petition-processing environment, and identify
potential improvements to the process. KPMG found that INS approved between
136,888 and 138,385 petitions--well over the limit of 115,000. KPMG also
found that the computer system that tracks the H-1B petitions was not
designed to count petitions against the annual limit. It found that the
computer system needs the capability to accumulate accurate and current H-1B
information, identify individuals who submit multiple petitions with slight
variations in biographical data and count them only once against the cap,
and, in general, support the generation of an accurate and timely H-1B
count. KPMG made a series of recommendations to INS to improve the accuracy
of the count, and INS is in the process of analyzing and incorporating these
suggestions.

Labor and NSF have taken several steps to improve the IT skills of the U.S.
workforce, as have IT employers. Labor and NSF's efforts have recently
provided millions of dollars in grants to institutions to train workers or
provide scholarships, respectively, to increase the number of American
workers with IT skills. While these efforts may over time help increase the
number of American workers with IT skills, their ultimate effect is unknown
given their recent start, their long-term focus and, in some cases, the lack
of data on specific IT skills needed. In the short term, some employers say
they train existing workers or new employees, while in the long term, they
reported that they are encouraging students to pursue IT careers.

In February 2000, Labor provided $12.4 million to nine grantees to train
employed and unemployed workers for high-skill occupations that are in
demand. In July 2000, Labor announced the second of three rounds of
demonstration grants and awarded $29.1 million to 12 grantees for training
American workers for high-skill jobs in areas where companies are facing
labor shortages.22 ACWIA, which mandated the skill grants, did not specify
what occupations the skill grants should target. As a result, Labor used as
a proxy those occupations for which employers requested H-1B workers. Given
that accurate data on H-1B approvals did not exist at that time, in its
August 16, 1999, Federal Register notice, Labor provided a list of the
occupations requested on LCAs from October 1, 1998, to May 31, 1999. The LCA
data showed health care and IT as the two industries most frequently using
the H-1B program. Labor officials noted that the actual number of
occupations shown by LCA data was flawed because, for example, some openings
certified on the LCAs for anticipated employment do not actually get
filled.23 Moreover, Labor's criteria for assessing the merits of each
proposal were based largely on characteristics of typical job training
program quality factors, such as service delivery strategy, target
population, and outcomes. Of the possible total points, 20 percent were
assigned for local needs for the first round and 15 percent were assigned
for the second round.

Despite these limitations, Labor's first round of 2-year grants was focused,
with one exception, on training in IT-related occupations (see app. II for a
listing of the grants). As a result, the grants are likely to contribute
U.S. workers with IT skills to the workforce. For those grants awarded under
the first round, grant recipients plan to train about 3,000 people in
IT-related skills. Under the second round of grants, most were also focused
on IT training; recipients plan to train about 2,500 in IT-related skills.
However, given remaining questions about the number and type of workers
needed, whether individuals will have the skills that employers need is
unknown.

In its first round of scholarship grant awards, NSF provided about $22.5
million to 114 academic institutions so that each institution could provide
approximately 40 scholarships per year over a 2-year period of up to $2,500
to low-income, academically talented students to help them pursue associate,
baccalaureate, or graduate-level degrees in fields such as computer science,
engineering, or mathematics (see app. II for examples of these grants).24
NSF also plans to award $25 million in the second round of awards25 and $24
million in another round of awards in 2001.26

According to NSF officials, NSF sought to fund those projects that best tied
the school's academic standards to the workplace by helping students make
the transition from school to work. Applicants were judged for

ï¿½ an infrastructure designed to help scholarship recipients graduate
(including, for example, academic support and mentoring);

ï¿½ a management and administration plan that is effective and clearly
articulated, which includes verification of scholarship candidates'
eligibility and evaluation of program outcomes; and

ï¿½ an education program of high quality, having external accreditation and
academic courses of study that are well defined, current, and intellectually
rigorous.

NSF officials believe the scholarships will lead to long-term outcomes such
as

ï¿½ improving education for students in the stated disciplines,

ï¿½ increasing retention of students to degree achievement,

ï¿½ improving professional development and employment and/or further higher
education placement of participating students, and

ï¿½ strengthening partnerships between institutions of higher education and
related employment sectors.

Given the long-term nature of the scholarships, it is difficult to know
precisely how many workers they will add to the U.S. IT workforce. First,
the actual amount of the individual scholarships is relatively small so that
these funds alone may not be sufficient to pay education costs. According to
NSF officials, the scholarships provide strong incentives to institutions to
retain and prepare students for IT-related occupations. It will be important
for the academic institutions to combine financial resources from a number
of sources to ensure that these scholarships make a difference. Second, it
is difficult to predict whether scholarship recipients will actually enter
IT-related occupations, because recipients' choice of majors does not always
correspond with their actual employment several years later. NSF officials
acknowledge that these individuals may not enter

the IT field, and also that students from other academic backgrounds may
enter IT occupations.

Term

IT employers we contacted said they have several choices to fill their many
IT vacancies in the short term--they can retrain existing workers, train new
workers, or recruit workers from the outside with the necessary skills. For
example, one employer retrained its workers whose jobs were being
eliminated, such as retraining a hardware engineer to be a software
engineer. According to the employer, the program costs about $3,500 per
worker for 3 months of training. Other employers established tuition
reimbursement programs to encourage their employees to obtain or improve
skills, and offered annual stipends, part-time employment, and full tuition
and fees while workers pursued a master's or doctoral degree in an
IT-related field. Other employers maintained skill inventory databases and
required employees to develop individual development plans in order to
update or obtain IT-related skills.27

Several employers said they train newly hired employees--even recent college
graduates--in order to give them the necessary IT skills, at a cost two
employers estimated to be about $10,000 for each new employee. Because of
high turnover, training for new workers is a potentially costly option, and
employers may be reluctant to provide training at significant cost for fear
that, once employees have received this training, they will leave.
Consequently, when employers do provide training, they may ask employees to
agree to stay for a period of time in exchange for the training provided--in
one case, it was for 1 year.

According to employers, the ideal short-term solution for filling IT
occupations is to obtain IT workers who already have the necessary skills.
To do this, employers use a variety of traditional recruiting strategies,
such as on-campus college recruiting and participation in job fairs.
Employers also use less-traditional recruiting strategies. For example, they
said a great deal of hiring occurs now over the Internet; employers either
post job vacancies and applicants respond, or employers recruit from web
sites where job seekers post resumes. Another method is "cold calling;" a
recruiter at a relatively small business noted that he sometimes calls a
company that he knows has the type of talent he needs and attempts to
recruit workers who answer the phone. Finally, employers also take advantage
of contract workers; a large employer that had over 300 contract personnel
from over 100 contract labor agencies reported that the company ensures that
all such contracts provide the option of eventually hiring the workers
directly.

However, IT employers we interviewed said that they cannot always find U.S.
workers with the necessary skills to fill all IT vacancies; as a result,
they may actively seek workers through the H-1B program. Employers may find
H-1B workers through their usual recruiting efforts, because workers who may
be already employed in the United States or on a student visa best meet the
employers' needs. The H-1B program offers a number of benefits, including
that workers will start their jobs with the requisite skills, and also that
if the H-1B worker is good, employers can sponsor the worker for permanent
U. S. residency. All of the employers we interviewed sponsored at least some
of their H-1B workers for permanent residency.28 These IT employers,
however, also said that there were disadvantages to using the H-1B program,
as it can cost over $2,000, including attorney and filing fees, to obtain a
worker with an H-1B visa, and it may take as long as 4 or 5 months before a
visa is approved.

Employers are also making some effort to improve the IT skills of American
workers over the long term. These efforts are predominately focused on
encouraging students to pursue careers that may be IT-related. For example,
several of the employers we contacted said they worked with universities to
improve the skills needed in the IT industry. One employer reviewed
university curriculums and served on panels and partnerships intended to
improve IT skills. Other employers provided mentors to students, computer
assistance, or computer equipment to elementary and secondary schools. To
assist in improving math and science curriculums, one employer funded a
program whereby local universities would work with four school districts to
help children stay in school and go to college. For the participating
universities, the employer provided scholarships and internships for
minority students majoring in engineering and computer science. This
particular program will cost $2.5 million over a 4-year period.

The H-1B visa program has helped employers fill specialty occupations on a
temporary basis; in fiscal year 1999 alone, over 130,000 individuals were
approved to work in the United States on an H-1B visa. The significant
growth in the number of such workers authorized to enter the country
indicates that the program is an important tool for hiring workers in
specialty occupations, especially IT-related occupations. However, as the
program currently operates, the goals of preventing abuse of the program and
providing efficient services to employers and workers are not being
achieved. Limited by the law, Labor's review of the LCA is perfunctory and
adds little assurance that labor conditions employers attest to actually
exist. Furthermore, the requirement that employers first file the LCA with
Labor before filing the same information with INS represents an extra,
time-consuming step that adds to H-1B processing time. Expanding Labor's
authority to question information on the LCA would provide additional
assurance that labor conditions are being met; however, this would likely
increase processing time substantially and the Congress has demonstrated its
desire that this process be handled quickly by establishing a short time
frame and limiting Labor's review. If the Congress wants to retain the
minimal review, it could consider eliminating altogether the separate filing
of the LCA with Labor and assigning the LCA review solely to INS. Because
Labor cannot now independently verify whether labor conditions will be met,
no current protections would be lost if INS were to subsume this process.
Moreover, because INS reviews much of the information on the LCA as a part
of its petition review, there would not be any additional resource needs for
INS adjudicators to perform this function. Finally, eliminating one review
step may also shorten the total approval time for H-1B workers, thereby
increasing employers' ability to get the workers they need in a timely
manner.29

Limitations governing Labor's ability to enforce H-1B requirements for
employers who have H-1B workers restrict Labor's ability to adequately
detect program noncompliance or abuse. Currently, unlike other labor laws it
is responsible for enforcing, Labor is able to initiate investigations to
address potential only if narrowly restricted circumstances are met. It has
no authority to subpoena records during an investigation to identify whether
employers are complying with the law, and it cannot conduct a baseline
survey to obtain a true understanding of employer compliance with the H-1B
program. Yet, Labor Inspector General reviews and other available evidence
suggests that program noncompliance or abuse by employers, after the H-1B
workers have been placed, exists and may be more prevalent than under other
laws where Labor has broader enforcement authority.

Finally, existing INS procedures do not give adequate assurance that program
noncompliance is being detected. INS' recent efforts to implement procedures
to standardize adjudicators' review steps are a positive step. However, INS
must continue to implement and revise the procedures and conduct training to
address the concerns raised by adjudicators, such as when and how they can
request additional documentation from employers. Additionally, there is no
requirement that adjudicators document their adherence to procedures or how
they exercised their discretion in assessing the sufficiency of
documentation when petitions are approved. The lack of easy access by
adjudicators to case-related information, as well as supervisory and
performance appraisal procedures that could discourage adjudicators from
denying petitions, can lead to incorrect approvals of employer petitions.

INS' efforts to upgrade its data systems and analyze and incorporate KPMG's
suggestions should prove beneficial to the entire H-1B visa process, but the
ultimate effectiveness of such computer upgrades and revisions will be
limited if INS does not include easy access by adjudicators to the
case-specific information they need to accurately assess the merit of
petitions. Finally, even with these improvements, unless INS has a
supervisory review and performance appraisal system based, at least in part,
on quality rather than quantity of review, it is not clear whether
adjudicators will have any incentive to carefully scrutinize the merit of
petitions.

The federal government and employers are making efforts to improve the IT
skill set of U.S. workers. While these efforts may help increase the number
of workers with IT skills, it is too early for two reasons to evaluate
whether the skill grants or the scholarships being funded by H-1B fees will
reduce the demand for H-1B workers. First, there continues to be a wide
range of views on the extent of need for IT workers and the kinds of skills
that are most important. Second, education and training programs, by their
nature, can be long-term remedies to labor needs. To be successful, these
programs will have to continually adjust their focus to the changing skill
needs of the rapidly growing IT industry. Better information now available
on the kinds of positions that H-1B workers are filling should help to
target their efforts.

Given the limited nature of Labor's review of LCAs, the Congress should
consider streamlining the H-1B approval process by eliminating the separate
requirement that employers first submit an LCA to Labor for certification.
Instead, the Congress could require employers to submit an LCA and the I-129
petition simultaneously to INS, which will continue to review and evaluate
the information contained on both the LCA and the petition.

If the Congress wished to broaden Labor's enforcement authority and improve
its ability to enforce relevant provisions in the H-1B law, it could
consider, at a minimum, giving Labor's WHD subpoena power to obtain
employers' records during investigations under the H-1B program. It could
also consider allowing Labor to perform baseline evaluations to determine
the extent of employers' compliance with H-1B requirements and conduct
subsequent targeted efforts to address suspected noncompliance or abuse.

To improve INS' ability to prevent H-1B visa abuse and better serve
customers, we recommend that the Attorney General direct the Commissioner of
INS to take the following steps:

ï¿½ expand upon INS' current efforts to standardize H-1B adjudication
procedures by (1) providing practical guidance to help adjudicators assess
the adequacy or sufficiency of documentation and determine when and how to
request additional documentation from employers, and (2) having adjudicators
document adherence to standard procedures when reviewing petitions;

ï¿½ provide easy access to case-specific information for adjudicators when
reviewing petitions as a part of the current upgrade of its computer system;
and

ï¿½ enhance existing supervisory review and performance appraisal systems so
that adjudicators are held accountable for the correct assessment of
petitions as well as for the quantity of reviews they complete.

The Departments of Labor and State, INS, and NSF commented on a draft of
this report (see apps. III-VI, respectively). Only Labor and INS provided
comments concerning our matters for congressional consideration or
recommendations.

Labor did not agree with our matter for congressional consideration
concerning the transfer of the LCA review; instead, it stated that
consideration should be given to how to improve the substantive nature of
the LCA review in a way that has minimal impact on timely processing. If the
Congress expands the LCA review, we agree that Labor is the most appropriate
agency to perform a more substantive review. However, unless the Congress
chooses to require a more substantive review, we believe it would be more
efficient to have the review done only once, by INS.

Labor agreed with our second matter for consideration to provide it broader
authority to enforce the H-1B program's requirements. Labor said that it has
long urged the Congress to reconsider and expand the narrow limits on its
enforcement authority.

INS said it was taking a number of actions to ensure consistent adjudicator
reviews, make information more accessible to adjudicators, and experiment
with measuring quality. However, it did not agree with our overall
conclusions or our recommendations to improve the consistency and
correctness of its H-1B decisions. Regarding our first recommendation
concerning guidance to help adjudicators assess the adequacy of information
and request additional documentation, INS agreed in principle with
standardizing the procedures used by adjudicators to review petitions.
However, it said its recently implemented procedures are sufficient to
address our recommendation. We agree that these procedures should help
adjudicators understand the steps they should follow to assess the merit of
petitions. However, it is too early to tell whether the procedures as they
currently exist will be sufficient to address the areas we identified where
adjudicators were uncertain about how to exercise their discretion. INS also
said that our recommendation to have adjudicators document their adherence
to standard procedures would detrimentally affect timely processing if
extended to approval cases. We acknowledge that an onerous requirement could
affect processing time, but believe that there are simpler ways for
adjudicators to demonstrate their adherence to standard procedures. For
example, a checklist with space for adjudicators to note areas of concern
and briefly indicate how they address those concerns would provide a
decision trail for a reviewer. Without some documentation about how cases
are decided it is more difficult for supervisors to assess whether approval
decisions were correct. Such decisions constitute the vast majority of
decisions and we believe it is important that INS supervisors assess both
approvals and denials to assure that the adjudication process is operating
effectively.

INS also disagreed with our recommendation to provide adjudicators with easy
access to case-specific information. INS said that adjudicators already have
access to information they need, such as laws, legal precedents, or
information obtained by investigators. However, in general, laws or
precedents are not specific to the case an adjudicator is reviewing because
they do not involve the same employers, workers, schools, or occupations.
Also, we found that communications between adjudicators and investigative
staff at the service centers may not be as routine or efficient as INS
headquarters envisioned. Furthermore, petition denial information is
available only to adjudicators in the same service center for 30 days, after
which the information is available nationally, but is not easily accessible,
and does not provide adjudicators the basis for the denial. Both
adjudicators and investigative staff said that unless the information is
automated and potential problems are flagged, adjudicators will not take
advantage of such information. INS did state, however, that the upgrades to
its computer systems will enhance access to some of this information.

INS said it values both the quality and timeliness of decisions by
adjudicators, but disagreed with our recommendation to enhance existing
supervisory review and performance appraisal systems so that adjudicators
are held accountable for the correct assessment of petitions as well as for
the quantity of reviews they complete. INS said that adding a quality review
process to its appraisal system would place an impossible burden on
supervisors because they would need to take statistically valid samples for
each adjudicator. Moreover, according to INS, the employee-to-supervisor
ratio has changed from six employees per supervisor to as many as 15
employees per supervisor, while petition decisions have significantly
increased. We disagree with INS' view that adding a quality review would
necessarily be unduly burdensome. As discussed, we believe that INS' current
supervisory review and appraisal process emphasizes the quantity of reviews
over the quality, and that more balance is needed between the two
objectives. Headquarters officials said that service centers are, in fact,
already using a variety of techniques to check the quality of adjudication.
Although we did not observe these efforts at the service centers we visited,
INS said that efforts range from teams randomly selecting cases and
determining the permissible range of decisions before providing the cases to
an adjudicator, to supervisors reviewing a random sample of 10 cases for
each adjudicator. Moreover, supervisors already review denials and requests
for additional information. All of these efforts indicate that service
centers have found a need for some type of quality review and are
implementing some mechanisms. We believe efforts such as these need to be
established agency-wide so that quality of adjudication receives the proper
degree of attention.

All four agencies provided technical comments, which we incorporated as
appropriate.

We are sending copies of this report to the Honorable Alexis M. Herman,
Secretary of Labor; the Honorable Janet Reno, Attorney General; the
Honorable Doris Meissner, Commissioner, Immigration and Naturalization
Service; the Honorable Madeleine Albright, Secretary of State; the Honorable
Dr. Rita R. Colwell, Director, National Science Foundation; and other
interested parties. We will also make copies available to others upon
request.

If you or your staff have any questions about this report, please call me on
(202) 512-7215. Other GAO contacts and staff acknowledgments are listed in
app. VII.

Marnie S. Shaul
Associate Director
Education, Workforce, and
Income Security Issues

Scope and Methodology

To obtain information on the jobs H-1B workers are filling in the United
States and the workers' characteristics, we relied primarily on data
collected and supplied by the Immigration and Naturalization Service
(INS)--most of it from a sample of petitions filed for fiscal year 1999. INS
randomly sampled 1,100 petitions processed at each of the four service
centers from a universe of petitions approved during the period from May 11,
1998, through July 31, 1999, and projected the results to the universe of
134,411 H-1B petitions approved during that time, weighting them to reflect
the variations in numbers of petitions processed among the four centers. The
universe included only applications for new employment (not renewals or
changes of employers)--those that potentially would have been recorded
against the fiscal year 1999 limit. INS employees reviewed the petitions in
the sample and accompanying documents at the INS records center in
Harrisonburg, Virginia, and recorded data from those files. We analyzed
these data further to, for example, make comparisons between information
technology (IT) and non-IT workers. INS also provided results from an
analysis of all petitions approved for work beginning during the first 5
months of fiscal year 2000. We did not independently verify the data
collected by INS in either the 1999 sample or the 2000 data. However, we did
compare the results of our analysis of the INS data with information
published by INS.

To assess the adequacy of the H-1B visa program's implementation and
enforcement, we interviewed officials at the headquarters offices of Labor,
INS, and State to understand the policies and procedures of the program. To
understand how employers and nonimmigrants must proceed through H-1B visa
approval, we visited three of INS' four service centers (in Laguna Niguel,
California; Dallas, Texas; and St. Albans, Vermont) and three Labor regional
offices (in San Francisco, California; Boston, Massachusetts; and Dallas,
Texas), contacted officials at the fourth INS service center (in Lincoln,
Nebraska), and obtained information electronically from eight State consular
offices with high H-1B visa workloads. We discussed how prevailing wage
rates are calculated with an official at one state employment service
agency. In addition, we obtained views on H-1B procedures from IT employers
who use the program, and met with associations representing American
employer and employee groups, including representatives from higher
education.

To identify and obtain information about efforts under way to improve IT
skills in the American workforce, we met with officials at the National
Science Foundation and Labor to obtain information on the training and
scholarship programs funded with H-1B visa fees. We also contacted 13
employers nationwide, representing companies that significantly invest in
software development or provide IT services, and one university. These
employers were selected judgmentally based on several factors and included
employers who employed H-1B workers, geographic representation, and
different company sizes. We also contacted associations such as the
Information Technology Association of America, the National Association of
Manufacturers, and the American Immigration Lawyers Association. In
addition, we contacted groups representing employees such as the AFL-CIO and
other unions, and the Institute of Electrical and Electronics Engineers,
Inc.

Labor's Technical Skill Grants and Examples of NSF Grants Awarded

The technical skill grants that Labor awarded as of February 2000 are listed
in table 2 below.

          Awardees              Award    Project emphasis and target group
                               amount
                                         Create a sustainable network of
                                         training providers to train and
 Regional Employment Board of            upgrade the technical skills of
 Hampden County, Inc.,        $1.5       130 employed and 80 unemployed
 Springfield, Mass.           million    individuals for highly skilled
                                         jobs in the information and
                                         telecommunications technology
                                         industry.
                                         Bring together a consortium of
                                         partners to provide high-level
                                         technical skills training over 2
                                         years to 200 individuals from the
                                         poorest neighborhoods in the
 NOVA Private Industry        $1.3       Silicon Valley with predominantly
 Council, Sunnyvale, Calif.   million    Hispanic, African-American, and
                                         Pacific Islander populations. The
                                         grant will also target low-income,
                                         multiethnic adults and older youth
                                         (18-24), dislocated workers, and
                                         incumbent workers.
                                         Build on a project to provide
                                         training to a wide range of
                                         participants in five "H-1B
                                         technical skill areas" that are in
                                         short supply in Pima County:
 Pima County Community                   health, IT (up to 180
 Services Department, Tucson, $1.5       participants), education,
 Ariz.                        million    electrical and electronics, and
                                         accounting and management. The
                                         grantee is committed to training
                                         single parents and women, with a
                                         focus on nontraditional fields for
                                         women.
                                         Develop an information technology
                                         worker training model that meets
                                         the needs of business and includes
                                         both entry into the IT industry
                                         and paths to career advancement.
                                         About 425 employees of
 The City of Chicago, Ill.    $1.5       participating companies will be
                              million
                                         selected, based upon prerequisite
                                         training and job performance, to
                                         train for H-1B designated
                                         positions, with 200 openings
                                         created through promotion and
                                         training.
                                         Develop new building blocks for IT
                                         training, geared to several target
                                         populations at various stages on
                                         the career ladder, and provide
                                         training for up to 500 people.
 Seattle-King County Private             Target groups for training are
 Industry Council, Seattle,   $1.5       unemployed workers who require
 Wash.                        million    additional skills to gain
                                         employment in IT professions and
                                         currently employed workers who
                                         wish to advance in their
                                         professions or change career
                                         paths.
                                         As a partner with two employers,
 The Workplace, Inc.,         $1.5       provide IT-related skills training
 Bridgeport, Conn.            million    to about 540 workers. Target
                                         population includes both
                                         unemployed and employed workers.
                                         Address needs of area employers
                                         for nurses at all levels and
                                         especially for the highest-skilled
 Philadelphia Workforce                  nurses--registered nurses (50) and
 Development Corp., Inc.,     $0.6       licensed practical nurses (30). In
 Philadelphia, Pa.            million    addition, 200 will be trained as
                                         nurses' aides. Training will
                                         target incumbent workers, low-wage
                                         workers, younger workers, and the
                                         unemployed.
                                         Implement job training and career
                                         development program to increase
                                         companies' ability to find and
                                         retain skilled workers in the
 New Hampshire Job Training   $1.5       state. One important innovation is
 Council, Concord, N.H.       million    the use of a newly developed
                                         bachelor's degree program in
                                         information sciences. The grant
                                         will target up to 320 unemployed
                                         and incumbent workers around the
                                         state.
                                         Recruit, assess, train, and place
                                         588 participants into jobs in the
                                         telecommunications and IT fields.
                                         Employed and incumbent workers in
 Prince George's Workforce               the East Bay area of Northern
 Services Corp., Landover,    $1.5       California and the Washington,
 Md.                          million    D.C., metropolitan area will be
                                         targeted, with emphasis on a
                                         nontraditional information
                                         technology workforce which
                                         includes minorities, women, and
                                         handicapped workers.

Rensselaer Polytechnical Institute plans to provide scholarship
opportunities to 40 participants, and has the goal of providing sufficient
support services to graduate 95 percent of these participants and implement
a series of workshops to engage participants in research and community
service activities relevant to their degree program.

Houston Community College in Texas plans to provide scholarships, curriculum
enrichment, student support services, and summer internships. It will
recruit 40 students into an Associate of Science and Associate of Applied
Science scholarship program and has a goal of retaining at least 75 percent
of the participants until they complete a degree and transfer to a higher
degree program.

The University of Texas at El Paso plans to offer scholarships to 22
upper-division undergraduate students in computer science, engineering, and
mathematics; 11 master's level students in those programs; and 11 doctoral
students in computer engineering, environmental science and engineering, and
materials science and engineering who are at the dissertation-writing stage
of their graduate careers. These students will also receive mentoring from
faculty who combine outstanding teaching with research programs funded
through federal, state, and corporate sources.

Comments From the Department of Labor

Comments From the Department of State

Comments From the Immigration and Naturalization Service

Comments From the National Science Foundation

GAO Contacts and Staff Acknowledgments

Lori Rectanus, (202) 512-9847

Carol L. Patey, (617) 565-7575

In addition to those named above, Betty S. Clark, J. William Hansbury, Jr.,
Lawrence J. Horinko, John G. Smale, Jr., and Joan K. Vogel made important
contributions to this report.

(205503)

Table 1: H-1B Investigations, Violations, and Back Wages Due 22

Table 2: Labor's Technical Skill Grants Awarded, February 2000 42

Figure 1: New H-1B Nonimmigrants Approved, Fiscal Year
1992-Fiscal Year 1999 8

Figure 2: Summary of H-1B Visa Approval Process 9

Figure 3: Occupations for New H-1B Visas, Fiscal Year 1999 15

Figure 4: Age Distribution of Workers Approved for H-1B Visas,
Fiscal Year 1999 16

Figure 5: Countries of Birth for New H-1B Visa Approvals, Fiscal
Year 1999 17
  

1. Under the H-1B program, specialty occupations are those requiring
theoretical and practical application of a body of specialized knowledge and
the attainment of a bachelor's or higher degree (or its equivalent) in the
specific specialty. These can be in a range of fields from architecture,
engineering, and mathematics to medicine, education, theology, and the arts.
Comparable wages are those being received by U.S. workers in similar
positions in the same area.

2. The Department of State also has a role in issuing the visas, as
discussed later.

3. The rest of the report focuses on only the specialty workers.

4. A consulting firm hired by INS estimated that INS approved between 21,888
and 23,385 more visas than were authorized for 1999, due to problems with
the computerized tracking system. Because the same system and approach were
used to count approvals in previous years, it is unknown whether INS
exceeded the authorized amount in previous years.

5. The following H-1B bills are pending before the Congress: S. 2045, H.R.
3983, and H.R 4227.

6. These procedures are followed whether an employer is requesting a visa
for a new worker or a renewal of a worker's existing H-1B visa. The H-1B
visa permits the worker to work only for the employer who originally filed
the request; if the worker changes employers, the new employer must obtain
new approval for the worker. Although there are limits on the number of
visas that can be approved each year, there are no limits on the number of
LCAs that can be submitted. According to Labor, it received over 300,000
LCAs in fiscal year 1999.

7. The employer may send the petition to the INS service center with
geographic jurisdiction for the work location, or may request, with
supporting justification, that one service center have sole jurisdiction for
processing all of its petitions, regardless of where the work site is
located. There are four INS service centers: Laguna Niguel, California;
Lincoln, Nebraska; Dallas, Texas; and St. Albans, Vermont.

8. See Information Technology: Assessment of the Department of Commerce's
Report on Workforce Demand and Supply (GAO/HEHS-98-106R, Mar. 20, 1998 );
U.S. Department of Commerce, Office of Technology Policy/Technology
Administration, The Digital Work Force: Building Infotech Skills at the
Speed of Innovation (Washington, D.C.: June 1999); and Peter Freeman and
William Aspray, The Supply of Information Technology Workers in the United
States (Washington, D.C.: Computing Research Association, 1999).

9. Labor noted that the H-1B fees build upon over $70 million in ongoing
efforts since 1998 to train workers in high-tech skills.

10. NSF will receive 4 percent for grants for math, engineering, or science
enrichment courses, and 4 percent for systemic reform efforts. Projected
funding for these two efforts is relatively small--$3 to $4 million annually
for 3 years. Because there is no time limit for expending the funds, NSF may
use the funds at a later date when there is a larger amount of funding
available for each effort, but has already expended some funds for several
small projects. Another 1.5 percent is available for INS to carry out duties
related to decreasing the processing time for petitions and to carry out
duties under ACWIA. Six percent is available to Labor for decreasing the
processing time for LCAs and investigating complaints.

11. INS randomly sampled visa petitions and supporting documents for workers
approved for H-1B status that count against the 1999 limit. Details about
the sampling and its results are included in app. 1.

12. Because our estimates are based on samples, they are subject to sampling
error. Our estimates have a 2 percentage point (or less) confidence interval
around each estimate. There is a 95 percent chance that the actual value
(whether dollars, ages, or percentage estimates) falls within that interval.

13. For example, the Institute of Electrical and Electronics Engineers' most
recent salary survey data indicated that one-quarter of its members reported
their area of technical competency as computer-related, with specialties
such as hardware development, software development, and network
administration.

14. One estimate of the current H-1B population in the United States is
360,000. See B. Lindsay Lowell, Ph.D., H-1B Temporary Workers: Estimating
the Population (Washington, D.C.: Georgetown University Institute for the
Study of International Migration, Apr. 17, 2000),
http:/www.ieeeusa.org/grassroots/immreform/h1breport.pdf (cited Apr. 24,
2000).

15. Difficulties meeting the 7-day timeframe led to the development, in
1999, of a facsimile process to speed up LCA processing. Machine
malfunctions, and heavy workloads for offices that still processed manually,
continued to lead to processing times in excess of 7 days; in February 2000
we found that, in two offices, LCAs were processed in 12 and 13 days.
According to Labor, as of May 2000, the facsimile process had matured, and
except during malfunctions or maintenance, regional offices are processing
the LCAs within the 7-day time period.

16. The source may not be an officer or employee of Labor unless the
information was lawfully obtained while conducting another investigation
under this or another act.

17. U.S. Department of Labor Inspector General, The Department of Labor's
Foreign Labor Certification Programs: The System Is Broken and Needs to Be
Fixed, Report No. 06-96-002-03-321 (Washington, D.C.: May 22, 1996).

18. INS headquarters officials said that adjudicators request additional
information on about 15 percent of nonimmigrant worker petitions they
review, based on unofficial local service center management reports. (These
data are not collected in any of INS' national systems.)

19. This effort involved INS service center staff screening petitions for
H-1B visas slated for issuance from the Chennai post, the most heavily used
consular post. State staff at the post then attempted to validate the
prospective workers' work experience or credentials.

20. This time frame represents the time between when the petition arrives at
the service center and when an adjudicator actually begins reviewing it.
Some petitions are decided on immediately; others take longer because they
are investigated further or the adjudicator requests more information.

21. According to INS, of the 2.5 million petitions and applications it
processes each year, about 500,000 are for nonimmigrant workers.

22. More funds are available; ACWIA permits these funds to remain available
until expended. Labor estimated nearly $80 million in fees received through
the H-1B visa program would be available to invest in high-skill training in
2000.

23. Labor included the same LCA data in the March 29, 2000, grant
solicitation for the second round of grants, even though INS by that time
had developed data on occupations for approved H-1B visa petitions. A Labor
official said they were aware of the new data but had not adequately
investigated it in time to include it in the grant solicitation.

24. Institutions were allowed to retain 9 percent of the grant for
administrative and academic support.

25. Proposals for the fiscal year 2000 funds were due to NSF by August 3,
2000.

26. NSF, and several other agencies, receive additional funds to develop
math and science curriculums for elementary and secondary students, intended
to improve students' skills needed in the future workforce. For information
on these other efforts, see Math and Science Education: Comprehensive
Information About Federally Funded Materials Not Available (GAO/HEHS-00-110,
July 12, 2000 ).

27. Approaches to attracting and retaining skilled information management
professionals are discussed in Executive Guide: Maximizing the Success of
Chief Information Officers--Learning from Leading Organizations
(GAO/AIMD-00-83, March 2000 ).

28. Although an employer may sponsor an H-1B worker for permanent residency,
this does not guarantee permanent residency. With increasing numbers of H-1B
workers, the annual numerical limitations on employment-based immigration
and per-country ceilings mean that growing queues and waiting periods face
H-1B workers seeking permanent residency in the United States.

29. There is a precedent for such streamlining. In 1997, we found similar
problems on another visa program--that for agricultural guest workers. We
found that INS was merely "rubber-stamping" visa applications, which
burdened the employer with additional paperwork and added time to the visa
application process. We recommended for that program that INS' role be
subsumed by Labor, which was performing the more substantive review; this
change was proposed in the Federal Register in 1999. See H-2A Agricultural
Guestworker Program: Changes Could Improve Services to Employers and Better
Protect Workers (GAO/HEHS-98-20, Dec. 31, 1997 ).
*** End of document. ***