Testimony of Stephen Yale-Loehr Adjunct Professor Cornell University Law School  

Hearing on Problems in the Current Employment Verification and Worksite Enforcement System 

Before the Subcommittee on Immigration, Citizenship, Refugees, Border Security, and International Law

Committee on the Judiciary U.S. House of Representatives Washington, D.C. April 24, 2007 

Madam Chairwoman and Distinguished Members of the Subcommittee:  

My name is Stephen Yale-Loehr. I teach immigration law at Cornell University Law School. I am also co-author of Immigration Law and Procedure, a 20-volume immigration law treatise. It is considered the standard reference work for U.S. immigration law. It has been cited by courts more than 400 times, including several times by the U.S. Supreme Court. I also chair the business immigration committee of the American Immigration Lawyers Association. I am testifying in my personal capacity.  

Thank you for inviting me to testify about the current employer sanctions program and recommendations to improve it through an electronic employment verification system (EEVS). I have been following and writing about this issue since 1986, when Congress first enacted employer sanctions.  

My testimony first provides a history of employer sanctions. I then discuss the Basic Pilot Program, which is the administration’s current effort to improve employer sanctions. Next, I discuss some systemic problems in the current employer sanctions program. I conclude with some recommendations.  

1. History of Employer Sanctions

            A. Pre-1986: The Texas Proviso  

Until 1986 no law made it illegal for an employer to hire an undocumented worker. In fact, in 1952, in passing legislation making it illegal for any American to “harbor” an undocumented individual, Congress stated that it was specifically not illegal to hire such an individual.[1] This came to be known as the “Texas Proviso.” It meant that employers were free to hire whomever they chose, without having to verify an individual’s eligibility to work. If an unauthorized worker was among the ranks of their employees, nobody questioned it and the employer was free to go on with business as usual without worry; it was simply up to the undocumented worker to avoid being caught by immigration authorities and deported.[2]  

            B. 1986: IRCA Enacts Employer Sanctions  

As Congress considered immigration reform in the early 1980s, debate over whether to enact employer sanctions was long and intense. The theory behind employer sanctions is twofold: (1) imposing penalties on employers of undocumented workers will deter the hiring of such noncitizens; and (2) because securing employment is the primary reason for illegal entry and residence, this will reduce incentives for illegal entry.

Some members of Congress argued that an employer sanctions program might place an undue burden on employers. Not only would all employers be subject to new paperwork obligations; employer sanctions also raised the specter that employers would have to become experts in immigration law to identify which categories of noncitizens were authorized to work.[3]  

Another area of concern was what kind of documentation would suffice to establish eligibility for employment. Congress was aware of the huge market that exists in fraudulent documents. Some proponents argued that a form of counterfeit-proof documentation ought to be devised to ensure the effectiveness of any employer sanctions program. The risk that this might lead to a “national identity card,” however, caused many members of Congress to shy away from such a requirement. Those opposed to an identity card won.

A third major concern was that employer sanctions would lead to discrimination against those who looked foreign or sounded foreign, and that existing fair employment laws would not provide a remedy. For example, Title VII of the Civil Rights Act of 1964[4] applied only to employers with fifteen or more full-time employees. Moreover, it barred national origin discrimination but not discrimination based solely on alienage.[5]

Congress addressed these competing concerns in the Immigration Reform and Control Act of 1986 (IRCA).[6] IRCA employed a three-pronged approach regarding the hiring of undocumented workers. First, it did away with the Texas Proviso and specifically prohibited employees from knowingly hiring undocumented workers.[7] Second, it required employers for the first time to verify, by use of the paper I-9 form still in use today, the identity and authorization to work of all their employees, including U.S. citizens.[8] For those individuals who failed to comply with or meet the new verification requirements, employers were required to refuse employment. Third, Congress included antidiscrimination provisions to prohibit employers from discriminating on the basis of national origin or citizenship status in hiring and firing employees.[9] Failure to comply with any of these provisions resulted in penalties being imposed, ranging from small monetary fines for first time, minor paperwork violations to criminal sanctions for repeat offenders, including jail time of up to six months.

To monitor whether employer sanctions would contribute to discriminatory practices, Congress asked the General Accounting Office (GAO) to prepare three annual reports on the employer sanction program’s implementation. Under the statute, the employer sanctions program would terminate if: (1) the final GAO report found that a “widespread pattern of discrimination” resulted from employer sanctions; and (2) Congress enacted a joint resolution stating that it approved the GAO findings.[10]

            C. 1990: GAO Finds Employer Sanctions Causes Discrimination

The GAO issued its final report in March 1990.[11] It was based on a survey of over 9,400 employers, which statistically projected to a universe of about 4.6 million employers. The GAO report found that the enactment of employer sanctions had created “a serious pattern of discrimination.” Overall, the GAO estimated that 19 percent of all employers began one or more discriminatory practices as a result of IRCA’s enactment.[12] The GAO report concluded that IRCA’s employer sanctions provisions failed to deter undocumented workers and increased discrimination against foreign-looking and -sounding workers because of: “1. lack of understanding of the law’s requirements, 2. confusion and uncertainty on the part of employers about how to determine employment eligibility, and 3. the prevalence of counterfeit and fraudulent documents that contributed to employer uncertainty over how to verify eligibility.”[13]

Despite the GAO report, Congress did not terminate employer sanctions. Over time, it became clear that employer sanctions was not working effectively. This was due in part to the conflicting interests that the sanctions try to satisfy and the lack of a mechanism to verify a worker’s actual identity and employment eligibility. Most importantly, the government never fully committed to seeing this new employee verification program through to fruition. The appropriate and necessary resources required to run the program were never devoted to it. As a direct result of this, and as the 1990 GAO report underscored, employers were simply unequipped to properly handle the large volume of fraudulent documents. At the time of IRCA’s implementation, 29 different types of documents were acceptable to verify work authorization and identity. With so many different documents allowed, this provided ample opportunity for fraud to take place. Employers, often having had little or no training in detecting fraudulent documents, were faced with the dilemma of either blindly accepting these documents or acting on a hunch and rejecting the documents but then facing penalties or lawsuits as a result of IRCA’s antidiscrimination provisions. In the end, fraud and discrimination took over the system.

            D. 1994: The Jordan Commission Proposes Government Verification

Employer sanctions created a paradox: a fairly high degree of supposed compliance but a relatively low degree of deterrence. Add the cost to businesses of paperwork and the documented increase in discrimination, and many people questioned the wisdom of continuing employer sanctions. One possible solution is to switch the burden of employment verification from employers to the government. The theory is that a government verification program could defeat the impact of fraudulent documents and also decrease discrimination by providing assurances to employers that the person they have hired is in fact authorized to work in the United States. The Commission on Immigration Reform, also known as the Jordan Commission after its chair, Rep. Barbara Jordan, was a leading proponent of testing such an approach.

 The Jordan Commission proposed having the government verify employment information by reviewing data from the Social Security Administration (SSA) and the immigration agency (at that time the Immigration and Naturalization Service (INS)).[14] Employers would submit employees’ social security numbers to a computerized registry system. The government would then verify that the number belonged to someone authorized to work.[15]

             E. 1996: IIRAIRA Changes

Congress tried to address problems in the employer sanctions regime as part of the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 (IIRAIRA).[16] The 1996 law, however, did not appropriate significant and necessary resources to improve workplace enforcement. Nor did it do much to improve the existing employer sanctions provisions. And even regarding document fraud, virtually the only measure taken by Congress in IIRAIRA was to slightly reduce the number of documents allowed for the purpose of verifying work authorization and identity, from 29 to 27. However, IIRAIRA did enact three pilot projects to strengthen and improve the employee verification process. Of these three projects, only one remains: the Basic Pilot Program.[17] The Basic Pilot Program is similar in concept to the Jordan Commission’s recommendation for a computerized registry system.

2. The Basic Pilot Program: Not Ready for Prime Time

The Basic Pilot Program, or the Employment Verification Pilot Program as U.S. Citizenship and Immigration Services (USCIS) now calls it, is a voluntary program whereby employers enter Form I-9 data (name; date of birth; Social Security number) into a computer within three days of an employee’s hire date.[18] This information is then compared with centralized databases at the SSA and the immigration agency (originally at the INS; now in the USCIS of the Department of Homeland Security (DHS)) to verify identity and citizenship. The data is then checked against a DHS database to verify employment eligibility.[19] If eligibility cannot be confirmed immediately, employers must notify their employees of the finding. The employees have the right to contest tentative nonconfirmation findings by contacting SSA or USCIS, as appropriate, to resolve any inaccuracies in their records. This contesting process is normally limited to 10 federal workdays. During this time, employers are not permitted to take any adverse actions against employees based on the tentative nonconfirmation finding. When employees contest their tentative nonconfirmation findings, USCIS informs their employers of the employees’ work-authorization status. When employees do not contest their findings within the allotted time, they receive final nonconfirmation findings. Employers are supposed to terminate employees in three circumstances: when employees indicate that they do not wish to contest the finding, when employees are found not to be work-authorized, or when employees receive final nonconfirmation findings.[20]

The Basic Pilot Program is not a perfect solution to the employer sanctions problem. The most fundamental problem remains the fact that for the system to work, DHS must run the identity information it receives against the SSA database, a database that is otherwise outside the Basic Pilot Program system and that is not intended to be used for immigration purposes. While the verification process now runs relatively quickly for citizens, processing times will certainly increase if all employers are required to use the system.[21] The processing times will be even longer for noncitizens. Because the DHS and SSA databases are not fully integrated and often have difficulty communicating with each other in an efficient manner, the process can take two weeks or longer for noncitizens. This is simply too long for many employers to wait.

Furthermore, a high number of errors continue to be reported, slowing the process even more. It has been estimated that about 20 percent of all initial Basic Pilot Program entries are false-negatives, meaning that the applicant is originally thought to be not work eligible, but that a later review determines him or her to be work authorized.[22] Many of these initial errors occur for simple reasons, such as the transposition of a first and last name, or a name change because the worker recently married.  

With the Basic Pilot program running inefficiently and ineffectively on a voluntary basis, and with only 15,000 participants, expanding Basic Pilot in its current state and requiring participation by all 8.4 million employers would be a bureaucratic nightmare. Full scale implementation would also cost at least $11.7 billion per year, according to a 2002 study.[23]

3. Systemic Problems with the Current Employer Sanctions Regime  

Several reasons exist for the failure of the current employer sanctions regime:  

•     First, the political compromise that formed the basis of employer sanctions in 1986 foundered on the economic reality of the continuing need for workers and the inability of our immigration system to provide them legally. That reality continues today.  

•     Second, as stated above, employer sanctions has created a paradox: a fairly high degree of compliance but a relatively low degree of deterrence. Employers are checking workers’ papers, as they are supposed to. But many of the verification documents may be fraudulent, or belong to a different person than the one who is presenting them. Until we solve the problem of fraudulent documents, employer sanctions will not work.  

•     Third, employer sanctions enforcement has not been consistent.  

These themes are fleshed out below.  

As recently as 2005, a Government Accountability Office study said that correcting the many problems out of the employee verification system remains the single biggest step in curtailing illegal migration and unauthorized employment.[24] But part of the problem is that we lack enough legal channels for foreign workers to come to the United States legally. Any comprehensive immigration reform bill must include temporary worker provisions if employer sanctions is to work.  

Another part of the problem is inconsistent enforcement of employer sanctions. As early as 1991, for example, Doris Meissner, former INS Commissioner, and Robert Bach, former head of INS policy and planning, warned that “evidence is building that the early effort among employers to comply in response to publicity about the new law and wide-ranging INS contacts is dissolving into complacency as employers experience the low probability of an actual INS visit.”[25] Their concerns proved valid. Government audits of employers to measure compliance with employer sanctions peaked at almost 10,000 in 1990, and then fell 77 percent to less than 2,200 in fiscal year (FY) 2003.[26] Notices of intent to fine companies for employer sanctions violations also dropped precipitously. From FY 1992 to FY 2004, notices of intent to fine fell 98 percent, from 1461 notices in 1992 down to just three in 2004.[27] Reflecting these trends, in 1999 the INS placed employer sanctions enforcement last in a list of five interior enforcement priorities.[28]  

Employer sanctions enforcement has increased over the last two years. U.S. Immigration and Customs Enforcement (ICE) has begun what DHS Secretary Michael Chertoff calls a “strategic shift” in enforcement by focusing on employers that knowingly or recklessly hire illegal immigrants. Those employers often face criminal charges — including money-laundering charges — and seizure of assets rather than administrative fines. 

“We found that [administrative] fines were not an effective deterrent,” Julie L. Myers, assistant secretary for ICE, told the New York Times. “Employers treated them as part of the cost of doing business.”[29] While the former INS brought 25 criminal charges against employers in 2002, ICE arrested 716 employers in 2006.[30]  

“Companies that utilize cheap, illegal alien labor as a business model should be on notice. ICE is dramatically enhancing its enforcement efforts against employers that knowingly employ illegal aliens,” said Ms. Myers in mid-2006. “Criminal indictments . . . are the future of worksite enforcement.”[31]  

That trend is continuing with some well-publicized raids. For example, just a few days before Labor Day last year, federal immigration agents descended on Stillmore, Georgia and surrounding areas just before midnight, entering homes and swarming the Crider chicken processing plant. Over three days, some 125 undocumented workers were rounded up and detained.[32]  

This effort to increase worksite enforcement became even more evident on December 12, 2006, when ICE agents raided Swift & Company production facilities located in six states.[33] This was despite the fact that Swift participated in the Basic Pilot Program. While I am not here today to address the human costs of the shortcomings of our immigration system, I would like to note here that the stepped-up enforcement efforts of ICE come at the steep if not incalculable cost of the lives of hard working families being torn apart. More quantifiable is the economic price of our failed employment verification system. Swift has stated that the raids, which displaced over 1,300 of their workers, would cost the company $30 million. A third of that will go to expenses tied to hiring incentives and work-retention efforts; the rest is tied to lost operating efficiencies.[34]  

Immigration raids nationwide have continued to increase this year, with recent raids in New Bedford, MA, Baltimore, MD and Santa Fe, NM, to name a few.[35] Enforcement of a broken system does not seem to be the just or economical approach that we, as a nation, should be taking. We are simply throwing good money after bad. Instead, we need to reform our employment verification system.  

4. Recommendations for a Workable Electronic Employment Verification System  

This much is clear: Some kind of workable, efficient, and accurate electronic employment verification system (EEVS) is necessary. The question remains, however: What steps need to be taken for this to be achieved?  

•     First and foremost, worksite enforcement must be part of a broader package of comprehensive immigration reform, which includes opening more—and more efficient—legal channels for foreign workers to enter the United States. Enforcement measures alone simply will not work. We must face economic reality and recognize that a labor shortage exists. For businesses to prosper and our country to remain competitive in the global economy, and to dissuade individuals from bypassing lengthy wait times and making unauthorized border crossings in search of jobs, employers need legal and efficient access to foreign workers.  

•     Second, the government must assist employers in making an employment verification system work, both through appropriate funding and by making all necessary resources, including money, technology, and additional training and manpower, available to them. The resources needed, while not fully identified at this point, assuredly will be great, and employers should bear some of the burden. But employers cannot be expected to comply with the law, verifying the identities and work authorization of tens of millions of potential employees fully on their own, without additional help from the government.  

•     Third, the government must make use of existing technology to create an efficient, workable EEVS. This includes continuing to implement and perfect biometric identification technology, possibly developing a biometric Social Security card, and simplifying the process for employers by allowing the use of a single swipe card containing information currently asked for on the I-9 form. The only documents that should be allowed to verify a noncitizen’s identity, immigration status, and eligibility to work are biometric Social Security cards, legal permanent resident (“green”) cards, and immigration work authorization cards.  

•     Fourth, once the problems with Basic Pilot or any other EEVS system are worked out--and they must be certified to have been worked out before full scale implementation—the system must require all employers to participate. Until that happens, however, the system must be phased in gradually, both to allow time for any necessary technological and/or efficiency fixes and to allow employers to acquire the tools necessary (biometric scanners, etc.) to implement and operate the system effectively. Given employers’ urgent needs for foreign workers, premature full scale implementation will result in skepticism in the program and an unwillingness to participate on the part of employers, who will choose to risk penalties and operate outside of the system.  

•     Fifth, DHS must establish some kind of entity to monitor the progress of new measures and the efficiency and accuracy of the program in general, and to help encourage employer participation.  

•     Sixth, any new employment verification system must be designed to protect privacy and ensure that the discrimination caused after IRCA’s enactment does not reoccur. To this end, the concerns of both employers and individual workers must be addressed. Any privacy violations or discrimination in the workplace must be quickly investigated and punished.  

•     Seventh, employer sanctions enforcement must be vigorous, consistent, and sustained. Employers initially complied with IRCA’s employer sanctions regime, in part because of vigorous enforcement across all industries. As enforcement waned in the 1990s, however, businesses began to worry less about employer sanctions compliance. Congress must appropriate enough money every year to ensure that employers comply with the law, and that ICE actively enforces it.  

It is relatively easy to state these goals. It is harder to know how to effectively implement them. For example, some people have advocated adding biometric information to Social Security cards so that they could be used as a reliable identity document for employment verification purposes.[36] The Social Security subcommittee of the House Ways and Means Committee held a hearing on this issue in March 2006.[37] That testimony deserves a careful reading. It is sobering. As Dr. Stephen T. Kent, chair of the National Research Council’s Committee on Authentication Technologies and Their Privacy Implications, testified, developing identity systems is much more complex than it initially appears:[38]  

Success . . . depends not only on the card technology we use but on all of the ways the system components have to work together. The high cost of fixing or even abandoning a system makes it essential that potential ramifications are explored very thoroughly prior to making decisions about design details and deployment of a system. . . . No method of ensuring that the person presenting the card is the proper owner can be completely reliable. A key decision for any system of this sort would be determining an acceptable threshold of false rejection and false acceptances, none of which are going to be zero in any practical technology. . . . In conclusion, . . . none of the issues raised by development and deployment of large scale identity systems are simple. The questions posed . . . should be carefully and thoroughly applied, not only from a privacy perspective but from a security, usability and effectiveness perspective as well.[39]  

Frederick G. Streckewald from the Social Security Administration testified at the same hearing that it would cost over $25 per card to issue a Social Security card with enhanced security features, such as biometric identifiers.[40] The SSA estimated that the cost of replacing Social Security cards for all 240 million Social Security cardholders would be approximately $9.5 billion.[41] That would not include the startup costs to buy the equipment needed to produce and issue such a card.   

Even assuming such the privacy, security, and cost issues could be worked out, other potential problems remain in using a biometric Social Security card or other national ID card. For example, such a card should not be issued to foreign nationals first. Otherwise, massive discrimination problems could result.  

Similarly, it is hard to know at what point any EEVS system will be reliable enough to impose on all employers. As numerous commentators have noted, even an error rate of just 1 percent would still translate into over a million people a year being erroneously disqualified or terminated from work.[42] Most of these would be U.S. citizens.  

We also need to have buy-in from employers and workers. Any system has to involve both groups to be effective. For that reason Congress should enact a provision to create an employer/worker advisory group to work with DHS in establishing an effective EEVS.[43]  

Finally, Congress should carefully consider the privacy implications of any electronic system to verify work eligibility. Employers will be forced to demand the required cards so it will become impossible to work in this country without carrying an identification card. This would be a fundamental policy change, because it would mandate ID as the cost of living and working in the United States. It would represent a fundamental reorientation of the relationship between the individual and government. Instead of being free to work, with the burden on the government to intercede where illegality is suspected, it would create an America where employees must seek the affirmative permission of government to work through the construction of a complex of databases and identity papers. And once that national identity infrastructure is created, privacy advocates worry that it would inevitably be expanded to many other purposes beyond preventing undocumented labor, including the routine monitoring and control of other activities.[44] Any EEVS system must have robust procedures to allow people to quickly fix errors.  

Many of these recommendations accord with provisions already in the Security Through Regularized Immigration and a Vibrant Economy Act of 2007 (STRIVE Act) (H.R. 1645). For example, the STRIVE Act would require ICE officials to spend at least 25 percent of their time on worksite enforcement.[45] That will help keep enforcement consistent and vigorous. Similarly, The STRIVE Act would expand existing antidiscrimination protections by applying them to the new EEVS set up under the bill.[46] The bill would make it an unfair immigration-related employment practice to terminate an individual based on a tentative nonconfirmation notice or to use the EEVS to screen an applicant before an offer of employment, among other things.[47] The STRIVE Act would also attempt to protect individuals’ privacy rights by storing only limited information in the EEVSA computer system.[48] The bill would also require employers to make sure others did not have access to the system.[49]  

Conclusion  

Employer sanctions is a multifaceted problem. It requires a multifaceted solution. It is like a three-legged stool. I call these the three Es--Enforcement, Evaluation, and Entry:  

•     Enforcement: We must have consistent and vigorous enforcement of our employer sanctions laws.  

•     Evaluation: We must have a mechanism of properly evaluating a person’s documents to know that they are not fraudulent and that they relate to the person presenting them. We must also continually evaluate any new employment verification system to make sure it is working properly and accurately, without creating adverse discrimination or privacy problems.  

•     Entry: We must create a temporary worker program large enough to allow most foreign workers to enter the United States legally. That will reduce the incentive to enter illegally.  

The three parts must be equally strong for employer sanctions to work. Failure to adequately address any of the three legs of this stool will mean that we will back here 20 years from now, discussing the same problem.  

Each of the three legs of the employer sanctions stool is a large problem itself. The overall problem cannot be corrected overnight. Congress and the American people need patience. Moreover, no one magic bullet exists for any of the legs. For example, the types of enforcement efforts may need to vary over time to keep up with new trends. Congress may need to try various pilot EEVS programs and then evaluate them. And more than one temporary worker program may need to be implemented.  

However, employer sanctions is a very important component of comprehensive immigration reform. It is perhaps the most important component, because it affects all Americans, not just immigrants. For that reason, it is imperative that we handle this issue carefully and thoughtfully. We may never be able to eliminate all undocumented workers, but we can work to make the problem manageable.