Immigration reform has many facets: It must pro­tect national security, uphold the rule of law, strengthen citizenship, and benefit the American economy.[1] The overall effect must be to reduce illegal immigration in to the United States. Although border security generally receives more attention,[2] serious enforcement of current laws prohibiting the employ­ment of illegals is also an important tool in an overall strategy to reduce illegal immigration.

The majority of persons who enter the U.S. ille­gally or unlawfully overstay temporary visas do so for purposes of employment. Employment of such indi­viduals has been illegal since 1986, although that law has never been seriously enforced. If access to employment were curtailed in accord with that law, many (probably even a large majority) of current ille­gal immigrants would leave the country voluntarily, and the number of future illegal entrants would be greatly reduced.

Since employment is the magnet that draws illegal immigrants into the U.S., it follows that the best way to reduce illegal immigration is to shrink the employ­ment magnet. To accomplish this without resorting to the method of routinely rounding up and deporting thousands of illegal workers only to have them return and obtain another readily available job, policy should focus on the businesses that hire illegal immigrants and let general employment rules rather than individ­ual arrests drive the reduction in illegal immigration.

The policy should be based on the principles of empowerment, deterrence,and information. It should empower honest employers by giving them the tools to determine quickly and accurately whether a new hire is an authorized worker. It should hold employers free from penalty if they inadvert­ently hire an illegal worker after following the pre­scribed procedures.

Further, the policy should empower honest employers by freeing them from the burden of com­peting with dishonest businesses that deliberately hire illegal workers. This means that it must deter dishonest employers who willfully employ unveri­fied and unlawful workers by imposing substantial penalties on the employers when such hiring occurs. For deterrence to work, however, both the government and employers must have timely and accurate information regarding new hires.

The most promising solution to this problem is a tool called E-Verify. A real-time, Web-based verifica­tion system run by the Department of Homeland Security (DHS) and the Social Security Administra­tion (SSA), E-Verify can determine with great accu­racy the authenticity of the personal information and credentials offered by new hires. In most cases, verification occurs almost instantly.

With the improvements suggested in this paper and those ongoing refinements to the existing pro­gram, E-Verify can be highly effective in reducing the employment of illegal immigrants. It is possible that an eventual mandatory use of E-Verify by all businesses could eliminate nearly all of the "on-the-books" employment of illegal immigrants in the U.S.-an estimated 4 million illegal workers.

Congress should reauthorize E-Verify as it cur­rently exists and work to expand its reach and effi­cacy significantly in recognition of the fact that the law prohibits employers from hiring illegal immi­grants and that the objective of E-Verify is to enforce that law.

This approach, combined with targeted enforce­ment and stiff penalties for those who do not com­ply with verification requirements, will significantly deter unlawful employment and thereby reduce the demand for illegal workers. That will, in turn, reduce the chief incentive of those seeking to enter the United States illegally.

Understanding the Employment of Illegal Immigrants

By most estimates, around 8 million illegal immi­grants work in the U.S. There are generally three means by which illegal aliens obtain employment.

1. Working "on the books" with a fictitious Social Security number. In this situation, the illegal alien is employed formally by a business, just as any other employee is. The employer withholds Social Security (FICA) taxes and files a W-2 tax form for the employee. The illegal employee presents identity documents to the employer showing that he is either a U.S. citizen or lawful immigrant entitled to work.
   
   These documents will contain a name, date of birth, Social Security number, and possibly a green card number, which are either partially or completely fictitious. The employer dutifully records this fictitious information on an official form called an I-9 and stores the form in a file cabinet. If the information on the I-9 were checked, it would immediately be found to be fraudulent; at present, however, there is no prac­tical means for the government to corroborate the information on the I-9.
2. Working "on the books" through identity fraud. In this situation, the illegal alien is also employed by a business just like any other employee. The employer withholds Social Secu­rity (FICA) taxes and files a W-2 tax form for the employee. The illegal employee presents iden­tity documents to the employer showing that he is either a U.S. citizen or lawful immigrant enti­tled to work.
   
   However, in this case, the name, date of birth, Social Security number and (in some cases) green card number on the documents corre­spond to the identity of a real U.S. citizen or lawful immigrant. To obtain employment, the illegal fraudulently assumes the identity of another real person. The employer records the fraudulent information on the I-9 and keeps the I-9 on file, but neither the employer nor the gov­ernment checks to determine whether the employee is the person he purports to be.
3. Working "off the books." In this situation, the employer deliberately conceals the employment of the illegal alien from the government. There is no public record of the employee, no W-4 with­holding form is filed, FICA taxes are not paid, and no W-2 statement is sent to the government. It is very unlikely that an I-9 form is completed or kept.

Of the millions of illegal immigrants in this country, the best evidence suggests that some 50 percent to 60 percent of this employment occurs "on the books."[3] It is unclear how much "on-the-books" employment of illegal aliens is done with fic­titious information and how much is done by iden­tity fraud.

To reduce illegal immigration, all three means of illegal employment must be addressed, but this need for a broad approach should not be used as an excuse to do nothing. Although it is true that reduc­ing "off-the-books" employment will be the most difficult task, that does not mean that the govern­ment should do little or nothing about the high lev­els of "on-the-books" illegal employment until it has devised a foolproof means of stopping underground employment as well. Proper policy should take the critical first step of controlling "on-the-books" employment of illegal aliens.

History of Employee Verification

In 1986, the Immigration Reform and Control Act (IRCA) granted amnesty and the right to U.S. citizenship to 3.1 million illegal aliens.[4] In exchange for this amnesty, Congress promised U.S. voters that the government would take effective measures to eliminate future illegal immigration. A major element of the promised policy was employ­ment security: measures designed to prevent or reduce significantly the future hiring of illegal immigrants within the U.S. But Congress has delib­erately failed to fulfill its 20-year-old promise.

IRCA made it unlawful for U.S. employers to knowingly hire illegal aliens and required employ­ers to examine worker documents to determine whether newly hired employees are eligible to work in the U.S. In practice, this system has proven to be ineffective because illegal aliens were able to obtain forged documents purporting to show that they were either lawful immigrants or U.S. citizens. Because employers were unwilling or unable to ver­ify the authenticity of these documents, the federal prohibition on the hiring of illegal aliens became nearly meaningless. Congress understands this situ­ation but has refused to take any action to correct it.

Ten years later, in 1996, Congress enacted the Illegal Immigration Reform and Immigrant Respon­sibility Act (IIIRA). As introduced, this legislation would have required universal verification of infor­mation on the I-9 forms.

Some business interests and the pro-illegal immigrant lobby blocked this effort. Instead, the act took a small tentative step toward enforcing the pro­hibition on the employment of illegals by creating a pilot Employment Eligibility Verification System (EEVS) to enable employers to determine the authenticity of documents presented by newly hired employees. The law required the EEVS (called Basic Pilot) to be implemented in a minimum of five of the seven states with the highest number of illegal aliens. These five states were California, Florida, Illi­nois, New York, and Texas. Nebraska was added to the program a few years later.

Participation in the EEVS by employers in these states was voluntary. In 2003, after six years of EEVS operation, Congress expanded the system to allow voluntary participation by employers in all 50 states.

At present, some 82,000 employers participate in the EEVS system (now called E-Verify) voluntar­ily. If current trends continue, over 5 million employees will be checked through E-Verify during 2008; this will represent slightly less than 10 per­cent of total new hires in the U.S. during the year.

In June of this year, the Bush Administration issued an executive order requiring all federal con­tractors to utilize E-Verify in hiring. This will increase future use of the E-Verify system to well above current levels.[5]

How E-Verify Works

E-Verify is a Web-based electronic verification system operated by the Department of Homeland Security and the Social Security Administration that allows employers to check instantly the authenticity of identity information provided by new employees. In general, the employer will use a personal com­puter to submit certain basic information concern­ing the employee (name, date of birth, Social Security number, etc.) to the government. The infor­mation is securely transmitted to DHS and SSA, and DHS checks the information to determine whether it corresponds to a U.S. citizen or to a work-eligible immigrant. In most cases, DHS can check and con­firm the employee information and transmit a defin­itive reply to the employer within seconds.

The E-Verify system thus provides a very quick and largely accurate check of the authenticity of information presented by newly hired employees. The system also provides a simple and rapid mech­anism to correct initial erroneous determinations. There are eight steps in the full E-Verify process.

1. Participating employers use E-Verify to confirm the legal status of newly hired employees. By law, employers are directed not to use E-Verify to screen job applicants.
2. Participating employers must electronically ver­ify the status of all newly hired workers within three days of hire, using information that an employee is already required to provide on the Form I-9. Employers enter information (name, date of birth, Social Security number, and citi­zenship status) into a simple form that is acces­sible on the Department of Homeland Security Web site.
3. DHS receives and checks the data. If the employee claims he is a U.S. citizen, his data are checked against the Social Security Administra­tion data files. If the employee claims he is a non-citizen lawful immigrant, his status is checked against the SSA database and then against the main DHS automated immigrant database at U.S. Citizenship and Immigration Services (USCIS). If the non-citizen employee data cannot be corrob­orated by the USCIS automated check, the case is referred to a USCIS immigration status verifier, who checks the employee's information against other DHS databases.
4. If the employee information submitted to E-Ver­ify is corroborated by the SSA database, by the USCIS automated database, or by the immigra­tion status verifier's review, DHS sends the employer an electronic positive confirmation notice certifying that the employee is an eligible worker. A full 94 percent of E-Verify submis­sions receive initial positive confirmations, most within three to five seconds of the query's being sent. The overwhelming majority of those who do not receive an initial positive confirmation are, in fact, unauthorized workers.[6]
5. If the information submitted by the employee does not match any information in the SSA and USCIS records, E-Verify automatically gives the employer the opportunity to double-check the submitted information for clerical errors. If cleri­cal errors are found, the employee's data can be resubmitted immediately, and a positive confir­mation can be received from DHS within sec­onds. If no clerical errors are found, or if the information still does not match any information in SSA and USCIS records, then E-Verify issues a tentative non-confirmation to the employee.
6. In the case of a tentative non-confirmation, the employee has eight working days to contest or clarify the non-confirmation by contacting a local SSA office (if he is a citizen) or a USCIS office (if he is a lawful immigrant). The employee then makes an appointment with the appropriate local office to determine the cause of the tentative non-confirmation. If the employee is in fact a citizen or a lawful, work-eligible immigrant, the tentative non-confirmation will, in almost all cases, be the result of a clerical error or obsolete information in the SSA or USCIS databases. These errors can be resolved quite simply. Information concerning naturalization status can be resolved by a toll-free phone call. Ninety-five percent of contested non-confirma­tions are resolved with a single phone call or appointment.
7. If the employee chooses not to contest the tenta­tive non-confirmation or has not provided infor­mation to alter the non-confirmation within eight working days, DHS sends a final non-con­firmation to the employer electronically.
8. After receipt of a final non-confirmation, the employer must either (a) discharge the em­ployee or (b) notify DHS that it plans to con­tinue employment. This allows employers to continue employment in situations where they are certain the non-confirmation is incorrect and will be rectified at some point.[7]

A survey conducted in 2004-2005 by the Office of the Inspector General (OIG) of the Social Security Administration found that employers using the EEVS were well satisfied with the system. All of the employers surveyed rated the program from "good" to "excellent." Ninety-eight percent said that they would continue to use the program voluntarily.[8]

Remarkable Accuracy: Confirmations and Non-Confirmations

In 2007, Westat, an influential private research firm located in Rockville, Maryland, conducted a formal evaluation of E-Verify and its predecessor, Basic Pilot, for the DHS. As part of this evaluation, Westat reviewed all 2.7 million employee submis­sions to Basic Pilot/E-Verify between October 2005 and March 2007. The Westat evaluation provides the most thorough information currently available on the operation of E-Verify.[9]

• During that period, 92 percent of submissions received an initial positive confirmation; around 1 percent of submissions received an initial ten­tative non-confirmation that was contested and converted into a final positive confirmation once information discrepancies were corrected; and around 7 percent of submissions resulted in final non-confirmations, nearly all resulting from initial tentative non-confirmations that were never contested.[10]
• Among all employees who were eventually found to be work-authorized, 99.4 percent received an initial positive confirmation, and 0.6 percent received an initial tentative non-confir­mation that was corrected by a brief visit to an SSA or USCIS office.[11]
• Among individuals found to be native-born cit­izens, 99.9 percent received an initial positive confirmation, and 0.1 percent received an initial tentative non-confirmation that was quickly corrected.[12]
• Among work-authorized, foreign-born individu­als, 97 percent received an initial positive confir­mation, and 3 percent received an initial tentative non-confirmation that was quickly corrected.[13]

Overall, the evaluation showed that E-Verify was very successful in distinguishing between autho­rized and unauthorized workers. As noted, 7 per­cent of submissions to E-Verify resulted in final non-confirmations; the Westat evaluation estimated that 95 percent of these final non-confirmations were, in fact, unauthorized workers seeking em­ployment.[14] If used widely, the system has the potential to block nearly all employment based on fictitious identities.

On the other hand, the Westat report estimated that 5 percent of final non-confirmations under the system may have been authorized workers. These misidentified workers represented about one of every 300 persons reviewed by E-Verify. The Westat evaluation reported no instances in which autho­rized workers who received a tentative non-con­firmation were unable to contest the ruling successfully and establish proper work authoriza­tion. Instead, the misidentified workers were indi­viduals who received an erroneous initial tentative non-confirmation but failed, for a variety of reasons, to contest that ruling.

In summary, the system did a very good job of identifying authorized and non-authorized work­ers. It also provided a process for correcting errone­ous initial findings. However, the system did generate a small number of erroneous final rulings when lawful workers failed to contest inaccurate initial rulings, an issue that will be discussed fur­ther below.

Accuracy of the Social Security Database Used in E-Verify

The Social Security Administration has assigned over 435 million Social Security numbers to individ­uals. When SSA assigns a Social Security number, it creates a Numident file, or master record, of the number and the individual to whom it is assigned.

When a newly hired employee is checked through E-Verify, the information provided by the employee is checked against information in the Numident files at SSA. Therefore, the accuracy and efficiency of E-Verify are contingent on the accuracy of the Numident files.

How accurate are the SSA Numident files? In 2004-2005, the Office of the Inspector General of SSA conducted an audit of the Numident files to assess their accuracy with respect to the operation of E-Verify.[15] The audit showed that Numident files were generally accurate. Although 4.1 percent of the Numident files were found to contain a data discrepancy, those discrepancies would rarely inconvenience lawful citizens and residents being checked by E-Verify, nor would they impede signif­icantly the ability of E-Verify to identify illegal immigrants seeking employment.

To understand better the results of the OIG audit, it is important to recognize the difference between "false negative" and "false positive" dis­crepancies in the Numident database. A "false neg­ative" discrepancy would result in a "tentative non-confirmation" being incorrectly issued for a citizen or legal resident lawfully permitted to work. That individual would need to go to the local SSA or USCIS office to correct the discrepancy. A "false pos­itive" discrepancy, on the other hand, could cause an unauthorized individual to receive an erroneous "positive confirmation" through E-verify.

While the OIG audit showed that 4.1 percent of Numident records contained a discrepancy, roughly two-thirds of the discrepancies were of the sort that might result in a "false positive" confirma­tion, while only one-third were the sort that would result in a "false negative" non-confirmation. More­over, the "false negative" discrepancies were simple clerical matters that could be corrected with very little inconvenience to lawful workers. The most common of these discrepancies were failure to change a woman's maiden name after marriage, clerical errors concerning date of birth in the Numident file, and failure to update information concerning naturalization.

The "false negative" error rates generated by Numident errors vary between groups. For native born citizens, the rate was 0.5 percent; for natural­ized citizens, the rate was 0.7 percent; and for indi­viduals listed as non-citizens in the Numident files, the rate was 10.4 percent.

Most cases in the third group involve immigrants who have been naturalized but have no record of that fact in their Numident files. Until very recently, out-of-date naturalization data in SSA data files were the largest source of erroneous tentative non-confirmations for E-Verify. If a worker reported he was a citizen but his SSA Numident file showed oth­erwise, E-Verify issued a tentative non-confirmation even if USCIS data would have shown that the indi­vidual was a citizen or otherwise work-eligible.

In May 2008, DHS implemented new proce­dures to fix this problem by checking individuals against both SSA and USCIS electronic records before issuing a tentative non-confirmation. With this fix, the effective "false negative" error rate for "non-citizen" Numident files should fall signifi­cantly. (Regrettably, USCIS does not have electronic records of naturalizations that occurred before 1995, so the current fix does not entirely eliminate this problem.)

In sum, the OIG analysis suggests that errors in the Numident files might cause about 1.6 percent of individuals to receive a "false negative" response from E-Verify. Since half of these "false negatives" would be caused by the naturalization data issue, the recent fix by DHS will reduce this rate signifi­cantly. The remaining individuals would receive a "tentative non-confirmation" from E-Verify that would require them to visit their local SSA office to update or correct the information in their Numi­dent files. Given the simple nature of the errors revealed in the OIG audit, correcting the Numident data in this manner should not prove difficult.

Moreover, the "false negative" error rate would shrink over time. The existing errors in the Numi­dent database have built up over decades, but each "false negative" error needs to be fixed only once. If use of E-Verify was required for all new hires, the bulk of existing "false negative" discrepancies would be purged from the Numident files within a few years, and the future, long-term "false negative" error rate would fall even lower.[16]

Reducing Erroneous Non-Confirmations in E-Verify

As noted, the Westat evaluation of E-Verify found that 6 percent to 7 percent of verification requests resulted in a final non-confirmation. The evaluation estimated that 95 percent of these final non-confir­mations represented illegal immigrants seeking unlawful employment. E-Verify therefore seems quite effective in identifying a high level of potential unlawful employment. However, the Westat analy­sis also estimated that around 5 percent of E-Verify's final non-confirmations may incorrectly identify lawful workers as unauthorized. The Westat figures thus suggest that as many as one lawful employee in 300 may receive an erroneous final confirmation from the present system.[17]

As also noted earlier, the Westat analysis identi­fied no cases in which the contest process failed to identify authorized workers properly; instead, all of the estimated erroneous non-confirmations were deemed to occur because employees receiving mis­taken tentative non-confirmations failed to contest the initial inaccurate ruling.

Westat determined that lawful workers may fail to contest an erroneous tentative non-conformation for four reasons.

• The newly hired individual may have found a better employment offer from another firm.
• The first employer may have continued to employ the worker irrespective of the non-con­firmation, thereby making a resolution of the issue unnecessary.
• The worker may have failed to understand the process for contesting the tentative non-confir­mation. (Lack of understanding of the verifica­tion process would undoubtedly become less frequent if E-Verify becomes broadly used.)
• Most important, the employer may have screened the job applicant through E-Verify improperly before hiring, received a tentative non-confirmation, elected not to hire the indi­vidual, and simply failed to inform the applicant that he had been tentatively identified as an unauthorized worker.

Government rules specify that E-Verify is to be used to check new employees immediately after they have been hired; employers are not supposed to use this system to pre-screen job applicants before hir­ing. In addition, employers are required to inform all workers of the results of their E-Verify screenings.

But not all employers follow these rules, and some may use E-Verify to screen applicants and turn down applications from persons with tentative non-conformations without ever informing them that a non-conformation has been received. If such improper use of E-Verify became widespread, it might be possible for a lawful work-eligible individ­ual with clerical errors in his SSA file to be turned down for a series of job openings without knowing the reason why. Even though this would occur very infrequently, it would be a real problem for the indi­vidual involved.

Fortunately, it is possible to reduce the frequency of such problems through two policies.

1. Reduce simple errors in the E-Verify system. Nearly all erroneous tentative non-confir­mations are the result of simple errors in the E-Verify databases. These commonly include misspelled names, clerical errors in date of birth, or missing date of birth.
   
   As noted, the most common source of error overall has been missing naturalization data in the SSA files; probably half of all erroneous non-confirmations have stemmed from this problem. DHS has recently taken steps to reduce this source of error significantly by checking both SSA and USCIS files before issuing a tentative non-confirmation. However, USCIS does not have data on naturalizations that occurred before 1995 in accessible electronic form. Con­verting these earlier naturalization data into an electronic form would further reduce this source of error.
   
   Another frequent source of erroneous non-con­firmations has been women applying for work using their married names while still being listed by their maiden names in SSA files. Strongly encouraging women to enter both maiden and married names as part of their input to E-Verify would reduce this source of error.
2. Give individuals opportunities to determine the accuracy of SSA data before applying for a job. Individuals should be given an easy oppor­tunity to confirm the accuracy of their personal SSA/USCIS files independently of the job applica­tion process. This change would enable all lawful prospective job seekers to check their Numident files for errors before applying for a job, thereby eliminating the infrequent but real prospect of being turned down for a job opening because of an erroneous tentative non-confirmation stem­ming from a clerical error.
   
   Under the proposed system, individuals could pay government-licensed contractors to corrobo­rate the accuracy of their personal SSA/USCIS files. The process would be the same as an ordi­nary E-Verify check but would not require the individual to be hired for a new job. Like normal employers, the contracting firms would not be able to access SSA files directly, but only to deter­mine whether the personal information submit­ted by a client matched the government's records.

Most faulty personal data in SSA files will need to be corrected at some future point irrespective of how the information is used in the E-Verify system. Under present conditions, an individual may not become aware of problems in his or her SSA file until applying for retirement benefits. But it is better to correct faulty SSA data sooner rather than later because individuals who are employed with faulty or mismatched SSA identity records may not get full credit for their employment for purposes of calcu­lating future benefits. While correcting errors in the SSA files may be slightly inconvenient, it is in the best interest of the individual to correct all problems as soon as possible.

Employer Perceptions of E-Verify

According to the Westat study, nearly all employers using E-Verify expressed high levels of satisfaction with the system. Ninety-nine percent reported that the computer interface was "user friendly" or "very user friendly." Ninety-six percent recently reported that using the system did not "overburden" their staff. Only 4 percent found the system "hard to use."[18]

On the other hand, 11 percent of employers reported that they encountered at least some diffi­culty or inconvenience in operating the system. The difficulties reported were generally modest. Com­monly mentioned difficulties included unavailabil­ity of the system during certain times, difficulties accessing the system, and training new staff to per­form verifications using the system. Some employ­ers also identified problems related to passwords. Others commented that local Social Security Administration representatives were not familiar with the Web Basic Pilot program, did not return their calls, or were unable to answer their ques­tions.[19] DHS continues to work to improve the operation of the system.

Costs to Business of Operating E-Verify

Opponents of E-Verify often charge that the system is too expensive for employers to use and that small firms, in particular, would face prohib­itive costs. Indeed, there is a cost to business from using E-Verify.

DHS has registered a number of "designated agents" who process E-Verify queries for other U.S. businesses on a fee-for-service basis. Any company that wishes to use E-Verify but does not want to learn to operate the system for itself can hire a des­ignated agent to do the clearances.

These firms, on average, charge between $2 and $15 per employee submission.[20] Moreover, a law requiring general use of E-Verify would create greater competition and economies of scale in E-Verify processing, reducing costs below the already low current levels. A firm that faces higher internal costs to operate E-Verify for its employees can con­tract out the process to a designated agent. Such contracting out is already a typical business prac­tice. For example, many small and moderate-size businesses already hire outside payroll service companies to perform their payroll and tax pay­ment functions.

The Office of Management and Budget estimated the costs to business of requiring federal contractors to use E-Verify for their employees. OMB estimated that firms would incur start-up and administrative costs of around $15 per vetted employee, primarily for the initial and recurring costs of training person­nel to use the system, and that the operational cost of actually processing individuals through this sys­tem (including the costs of dealing with temporary and final non-confirmations) would be around $6.70 per processed employee. Thus, the overall costs to business to administer and operate E-Verify would total about $22 for each employee checked.[21]OMB did not consider whether costs could be cut by contracting out with designated agents or other personnel service companies.

The Westat evaluation of Basic Pilot/E-Verify also found low employer costs. Firms surveyed in the study reported very low costs per employee to set up the E-Verify system and operate it over a year. For example:

• Firms with between 100 and 250 employees reported average setup costs and annual operat­ing expenses of $646, or around $4.00 per standing employee;
• Firms with 251 to 500 employees reported average costs of $746, or around $2.00 per employee; and
• Firms with between 501 and 1,000 employees reported average setup and annual operating costs of $473, or less than $1.00 per employee.[22]

Given these estimates, the costs of implementing E-Verify appear modest in comparison to the mag­nitude of the illegal immigration problem. There are roughly 63 million new hires in the U.S. each year. At an average cost of $10 per employee, compre­hensive use of E-Verify for all new hires would cost employers around $600 million per year, or around 1/200th of 1 percent of the economy. This sum equals approximately $6.00 for each household in the U.S. Of course, it is always important to be cau­tious when government adds new regulatory bur­dens to business, especially since these costs tend to be invisible to the taxpayer.

Moreover, the future costs to business of operat­ing E-Verify could be substantially reduced by the elimination of duplicative governmental processes. Employers are already required by law to provide nearly all of the worker information used in E-Verify to the National Directory of New Hires (NDNH), which is operated by the Office of Child Support Enforcement (OCSE) at the U.S. Department of Health and Human Services (HHS). Though E-Ver­ify is fast and provides real-time feedback to employers, the child support data collection system is very slow and provides little useful feedback.

If all employers used E-Verify, NDNH data col­lection would become redundant, and initial new-hires data could be collected through E-Verify and transmitted to OSCE, eliminating the need for duplicate data submissions from business. In this case, the extra cost for business to operate E-Verify could fall to zero, although there would still be the temporary cost of transitioning from NDNH to E-Verify.

Costs to Government of Expanding E-Verify

Currently, E-Verify is used to check less than one-tenth of total new hires in the U.S. If it were used for all new hires each year, the cost to government to operate the system would rise, with most of the added costs occurring in USCIS. This agency has estimated that a phased four-year expansion of E-Verify resulting in the ongoing coverage of all new hires in the fourth year would cost $765 million over the four-year period. Coverage of all new hires and current employees in four years would cost $838 million, or an average of $210 million per year.[23]

Expansion of E-Verify use would also result in added costs to SSA because extra staff would be needed to process the increase in contested "tenta­tive non-confirmations." SSA has estimated that a phased expansion to cover all current employees and all new hires annually would cost it $281 mil­lion over five years, or $55 million per year.[24]

In total, then, the cost to government of imple­menting universal E-Verify coverage would be around $1 billion over four years. This amounts to about $2.50 per U.S. household per year.

Additional government expenditures might be required to meet the costs of prosecuting employers who willfully disregarded the E-Verify system by employing workers found to be unauthorized; how­ever, fines on such employers could offset some or all of this enforcement cost.

Current Fines and Penalties for Hiring Illegal Aliens

It is illegal under current law to "hire, or to recruit or refer for a fee for employment" any indi­vidual who is known to be an authorized alien.[25] It is also unlawful to hire an individual without com­plying with the current employment verification system (i.e., examining identity documents and fill­ing out an I-9 form).[26] The law provides civil and criminal penalties for violations by employers, but these penalties are generally too modest to serve as a deterrent to unlawful hiring.

Civil Fines. Employers face civil fines of between $100 and $1,000 per employee for failure to examine documents and file I-9 forms.[27] They face steeper civil fines for deliberately employing persons known to be illegal aliens; these fines are from $250 to $2,000 per alien employee for the first offense and rise to from $3,000 to $10,000 per alien employee for repeat offenses. However, to impose these penalties, the government must prove that the employer knew the hired alien was unauthorized.[28]

Under the current system of civil fines, DHS nor­mally issues a "notice of intent to fine" (NOIF) to an employer found to be in violation and then negoti­ates an out-of-court settlement with the employer. A few cases are taken before administrative law judges. It is generally felt that these civil penalties are too low and enforced too infrequently to have much, if any, deterrent effect on employers.[29] Many employers simply treat the possibility of a small fine as a normal, minor cost of doing business.

Criminal Penalties. The law also provides for criminal penalties for a "pattern or practice" of hir­ing illegal aliens.[30] These fines may not exceed $3,000 per alien, but they may include imprison­ment of the employer for up to six months for the entire offense. Enforcement of these provisions requires an indictment and prosecution. Additional criminal charges relating to tax evasion, abetting of identity theft, and violations of labor law may also be apply in some cases.

In recent years, Immigration and Customs Enforcement (ICE) has downplayed the pursuit of civil fines as ineffective. Instead, it has sought con­victions and penalties for criminal violations, often working with a variety of other government agen­cies.[31] In 2007, ICE enforcement activities against employers resulted in over 850 criminal arrests and over $30 million in fines.[32]

While recent ICE enforcement actions send a message that the government has become more seri­ous about enforcing the law against employment of illegal aliens, these tactics necessarily require the collection of extensive criminal evidence and are very labor-intensive for the agencies involved. While commendable, ICE enforcement activities still affect only a minute fraction of employers who deliberately hire illegal aliens; moreover, it is possi­ble that current tactics may simply be too expensive to use as a broad-based enforcement strategy. In the long run, a comprehensive civil enforcement sys­tem, with higher fines and regular enforcement, might yield greater dividends than the current focus on criminal prosecution.

To be effective, an employment verification and enforcement system must provide clear and fair rules for employers to follow, a significant increase in the probability that violations will be detected, and major increases in the penalties for deliberate violations of verification procedures. An effective system also should not generate large, unnecessary enforcement costs for government. Ideally, a verifi­cation and enforcement system should serve as an effective deterrent for the overwhelming majority of employers and should provide fines and penalties for rogue employers that are sufficient to cover investigative and enforcement costs.

Targeted vs. Broad-Based Verification

Some suggest that employment of illegal immi­grants is focused in a few industries and that it is therefore unnecessary to establish employment ver­ification across the entire economy. Instead, a policy of targeted verification and enforcement in those particular industries will suffice to effect significant reductions in illegal immigration. This argument falls apart when one examines the facts.

Though illegal immigrants are disproportion­ately low-skilled, they are spread widely through the economy. Illegal immigrants appear to work most commonly in six industries (in order of descending scale): construction, restaurants, land­scaping, janitorial services, food processing, and hotels. In fact, only around half of illegals work in these industries.[33] Some 45 percent may be visa overstayers rather than unlawful border crossers,[34] and between 10 percent and 15 percent actually have a college degree. Thus, they are spread more widely through the economy than popular wisdom suggests.[35]

Moreover, employment verification through E-Verify is effective only against "on-the-books" employment. Such hiring is more likely to occur in higher-wage and higher-skill jobs and less likely to occur in low-wage fields such as construction and farm labor. This means that the "on-the-books" hir­ing of illegal immigrants is likely to be less concen­trated in a few low-wage industries than is "off-the-books" hiring.

For these reasons, employment verification laws should apply to all industries equally to serve as a substantial deterrent to illegal immigration, and law enforcement actions should be prioritized to address the most egregious situations first.

Principles of a Sound Verification and Enforcement Policy

Most illegal immigrants enter the U.S. with the intention of working. An effective policy that limits the hiring of illegals would likely cause many to leave the country and could reduce the future inflows of illegals. A meaningful policy to verify the legal status of workers and to enforce the existing prohibition on hiring illegals should be one of the primary tools at the center of any strategy to reduce illegal immigration.

A sound verification and enforcement policy should have at least 11 components.

1. Require universal employment verification. For 20 years, U.S. law has prohibited the hiring of illegal immigrants, but there has been no seri­ous effort to enforce this law. In order to curtail illegal immigration, this statute should finally be enforced by moving toward requiring all employers to use E-Verify to confirm the employment eligibility of all new hires and cur­rent employees.
2. Reauthorize E-Verify and provide adequate funding for implementation. The law autho­rizing E-Verify expires in March 2009. Congress should extend E-Verify as a voluntary program and provide the funding needed for its contin­ued expansion and improvement.
   
   Currently, less than 10 percent of the 63 million annual new hires in the U.S. are checked through E-Verify. While E-Verify should be expanded, attempting to expand E-Verify to cover all ongoing new hires immediately is impractical. The government should eventually allow checks by participating employers of any existing employees who have not yet undergone verification. E-Verify confirmation error rates should be closely monitored and continuously driven downward. DHS is already making progress in this area.
3. Improve government data to further reduce erroneous tentative non-confirmations and provide opportunities for individuals to review the accuracy of their personal data in government files. While the recently imple­mented procedures for confirming naturaliza­tion are a significant improvement, DHS should further reduce this source of error by upgrading its naturalization databases. DHS should further reduce E-Verify errors by encouraging individu­als to enter both maiden and married names of women into the system.
   
   Opportunities should also be created for indi­viduals to review the accuracy of SSA and USCIS data independently of the job application pro­cess. Individuals who follow the proper process but still lose employment or wages due to gov­ernment misfeasance might receive financial compensation from the government equivalent to lost wages.
4. Penalize employers who continue to employ workers who have failed verification. The E-Verify system will be ineffective if employers can continue to employ workers who have failed the authorization check. However, it is not neces­sary for DHS to perform expensive site visits to determine whether employers have complied with the law by discharging employees who received final non-confirmations.
   
   Instead, DHS and SSA should review quar­terly wage reporting data on an ongoing basis to identify employers who continue to employ workers after receipt of a final non-confirma­tion. Those employers should be informed that they are in apparent violation of the law and served with a notice of intent to fine. Increased funding should be provided to enable DHS and the Department of Justice to prosecute employers who deliberately violate the law, starting with those with the greatest number of violations.
5. Increase penalties, in law and in practice, for unlawful hiring. Most illegal immigration is employment-driven. To achieve significant re­ductions in illegal immigration, government must make it difficult for an illegal alien to work in the United States and easy for an employer to tell the difference between lawful and unlawful workers. To accomplish this, employers must be given an easy and unambiguous means to iden­tify unauthorized workers and face substantial penalties if they willfully violate verification standards. Only that combination will deter ille­gal employment.
   
   The deterrent effect of any law is a product of (1) the probability that an offense will be detected and punished and (2) the severity of the penalty. Where the probability of apprehension and the severity of penalty are low-as is the case with the current verification regime-the deterrence value of a law is minimal.
   
   From a practical point of view, increasing the probability of punishment requires consider­ation of the burden of proof required for govern­ment to obtain a conviction. With respect to employment verification, there are currently two types of offenses. The first relates to busi­nesses that employ workers without proper ver­ification. Under a universal E-Verify system, this would include employment without required verification after final non-confirmation[36] or entirely off the books. Proving this kind of con­duct in court is straightforward.
   
   The second type is willful violation of the rules prohibiting employment of individuals known to the employer to be unauthorized aliens.[37] This requires proving that a worker is in fact unauthorized to work and that the employer knew that the worker was unauthorized-an extremely heavy burden for government prose­cutors. To prosecute the second level of offense, the government would need to apprehend the worker in question and determine that he is, in fact, and illegal alien (rather than, for example, a native worker seeking to avoid child support or tax obligations). Given that there are around 8 million illegal immigrant workers in the U.S., general prosecutions based on this second level of evidence would be impossible.
   
   The problem is that the punishment for the first type of offense-which it is relatively easy for the government to prove in court-is so low that it does not deter illegal hiring. Most employers can dismiss the existing fines of $250 to $1,000 per employee as a minor cost of doing business. For example, a fine of $1,000 is roughly 5 percent of the average annual wage paid to illegal immigrant workers. This type of fine has virtually no deterrent effect.
   
   To be effective, these fines must be raised to the point where they actually deter businesses from hiring illegals and are sufficient to cover much of the cost of government investigation and enforcement. For example, the fine for unlaw­fully employing one worker might be set at $5,000, with higher fines for repeat offenses. The fine for unlawfully employing multiple workers might be set at $20,000 per worker (roughly equal to the annual wages of illegal immigrant workers). The fines for "off-the-books" employment, which would be more dif­ficult for government to detect and prosecute, should be even higher.
   
   In addition, tax law should be changed to bar employers from claiming the wages of employees hired without required verification as business expenses. DHS and SSA should share employ­ment verification data with the Internal Revenue Service to facilitate enforcement of this rule.
6. Facilitate information sharing between DHS and SSA. SSA currently has information on spe­cific employers who have large number of employ­ees with erroneous Social Security numbers or other mismatched Social Security data. For exam­ple, in 2002, over 70 employers had more than 5,000 mismatched employees apiece.[38]
   
   A significant number of these employees are probably illegal immigrants, and their employ­ers should be prime candidates for enforcement action by ICE. At present, however, SSA is barred from sharing this information with DHS. This barrier should be removed. Enforcement focused on apparent heavy offenders is an important and efficient tool in combating illegal immigration and an essential part of broad-based employment verification.
7. Issue clarifying letters to employers regard­ing Social Security mismatch notifications. Each year, SSA issues letters to employers who have large numbers of employees with false or mismatched Social Security numbers. The employers are encouraged to take steps to rectify these cases.
   
   Recently, DHS sought to issue letters to employ­ers notifying them that failure to rectify the mis­matched Social Security numbers of employees could, in some circumstances, be regarded as knowing and willful employment of illegal immigrants. A federal district judge in California blocked the issuance of these letters, citing a number of technicalities. The Bush Administra­tion has now resolved these technical issues and should issue the letters. This would be an important interim step in employment enforce­ment, and it is one that can be taken immedi­ately because it does not require new legislation.[39]
8. Do not restrict state efforts to limit the em­ployment of illegal aliens. State governments such as Arizona's are currently leading the national ef­fort to limit the employment of illegals. Federal legislation should not impede these efforts.[40]
9. Establish supplemental procedures to pre­vent employment by means of identity fraud. One apparent shortcoming of E-Verify is that it cannot prevent identity fraud. The E-Verify sys­tem can easily detect an illegal alien who seeks employment using a fictitious Social Security number and date of birth. In this situation, a non-confirmation can be issued and the employment prevented.
   
   However, E-Verify cannot detect an illegal seek­ing employment by identity fraud-that is, by using a legitimate name, Social Security number, and date of birth copied from another real per­son.[41] E-Verify can determine only whether reported data are legitimate, not whether the prospective employee presenting the informa­tion is the person he purports to be.
   
   The way to solve this dilemma is through the simple procedure of notifying the real individual whenever a potential identity theft has occurred. Specifically:
   
   SSA would scan its wage database to identify individuals who held two or more jobs simul­taneously over an extended period, received Social Security benefits while working, or were employed while under age 16.[42]
   
   SSA would send a letter to the individual noti­fying the individual that a potential identity theft may have occurred. (Alternatively, it might be sufficient to target cases in which identity fraud seemed most likely, such as cases involving individuals with many simul­taneous jobs, individuals working in two or more widely distant locations, retirees, or children working in construction.) The letter would be sent to the individual's home address as reported to SSA on the individual's W-2 tax form.
   
   The notice would identify each job apparently held by the individual according to SSA records. The letter would inform the individ­ual that if he actually worked at each job listed, there was no problem and the letter should be ignored; however, if any of the jobs listed were work he had never performed, the letter would ask him to notify DHS or SSA through specified telephone number or Web site.[43]
   
   When SSA was contacted by an individual who stated that he was on record as holding a job he never performed, SSA would identify the suspect job site. SSA would share the information with DHS, notify the employer at that site that a potential identity fraud had occurred, and issue a tentative non-confirma­tion for the suspect employee. The employee would then have eight working days to con­test the non-confirmation at his local SSA office and demonstrate the authenticity of his claims of identity. If the individual failed to contest the non-confirmation within eight days and offer evidence of his identity, a final non-confirmation would be issued, and the employer would be required to fire the employee or provide SSA with a rationale for retaining him.
   
   This system offers a simple way to block most em­ployment through identity fraud. Most Americans would be eager to notify the government if they felt their identity had been used for fraudulent purposes. Conversely, few persons engaging in identity fraud would be willing to risk con­testing the tentative non-confirmation issued under this system.[44]
   
   In addition to the above steps, SSA and DHS should be empowered to audit reported earnings data to pinpoint cases in which identity fraud was most likely. These might include, for exam­ple, individuals apparently working simulta­neously in a very large number of different jobs or individuals working simultaneously at two widely separated locations. Special effort could be taken to investigate such cases.
10. Establish supplemental procedures to reduce "off-the-books" employment by illegal aliens. As noted, it appears that roughly half of illegal immigrant workers are employed off the books. While on-the-books employment of unautho­rized workers can be detected and blocked en masse through automated systems, detection of off-the-books hiring is far more difficult. Such detection must inherently involve piecemeal labor-intensive investigations, and the task is made more difficult by the fact that off-the-books employers in fields such as restaurants, construction, and agriculture are likely to be small-scale.
11. Incorporate the current new-hire data collec­tion for child support into E-Verify.[45] E-Verify and the new-hire data system run by the Office of Child Support Enforcement at HHS collect very similar information on newly hired employees; both check employee information against the SSA databases for accuracy.
    
    However, there are major differences between the two systems: E-Verify is designed to give accurate, immediate responses to employers, usually within a few seconds. By contrast, the new-hire system is intended to locate workers for purposes of collecting child support pay­ments. It is not designed to provide quick turn­around. The new-hire system is run by state child support agencies or other state offices that have no interest in detecting illegal immigra­tion or conducting employment verification. Notifying employers of Social Security number mismatches is not a central function of the new-hire data system. Feedback to employers, when it does occur, can take weeks or longer, and many states do not notify employers at all.
    
    The two systems collect similar data and there­fore could be integrated into a single data collection process, sparing employers the unnecessary time and effort involved in dealing with duplicated entries. However, since DHS needs immediate real-time responses for employment verification and OCSE does not, E-Verify should serve as the initial portal for data collection for both systems. Data from E-Verify would be corroborated by the SSA data­bases and then forwarded to OCSE for entry in its National Directory of New Hires. This should fulfill the needs of both DHS and OCSE. By contrast, any attempt to run DHS employ­ment verification through the slow, antiquated, and cumbersome child support system would be unworkable. Such a policy would set employment verification back by a decade.[46]

Conclusion

It is imperative for the nation to re-establish the rule of law concerning immigration. Our nation must gain control of U.S. borders and reduce the number of illegal aliens in the United States.

However, the choice between mass roundups and amnesty presented by some politicians is a false dichotomy. A third alternative exists: By seriously enforcing the laws against illegal immigration in the U.S., the government can remove the incentives for illegals to enter and remain in this country, thereby causing the bulk of them to return home and sharply reducing future inflows.

E-Verify has proven to be quite effective, and efforts to improve the system itself are ongoing. E-Verify is the most promising employment verifica­tion system in existence in the United States, and it should be continued.

Businesses must be able to distinguish quickly and accurately between lawful and unlawful employees. The government must have ready and instant access to information so that it can know when an unauthorized worker has been hired and can assist the employer in complying with the law. Businesses that deliberately seek to evade the law must face penalties that are substantial and suffi­cient to deter them from knowingly and willfully hiring unauthorized workers.

It is time for Congress to keep its promise. Under current law, employers are prohibited from hiring illegal immigrants. E-Verify has shown itself to be a highly effective screening and enforcement tool. The government continues to improve and expand the program, and error rates are at historically low rates, especially for a federal government program.

In moving forward, the objective is to achieve the goal set by Congress 20 years ago. Our political leaders cannot hide behind the "it needs further improvement" mantra forever, because doing so is the functional equivalent of not enforcing the prohi­bition against hiring illegals in the first place. With the improvements suggested in this paper and those that DHS has been implementing, we are fast approaching the time when past promises can become the reality of a system that enforces employ­ment immigration laws across the country.

-Robert Rector is Senior Research Fellow in Domestic Policy Studies at The Heritage Foundation.

Appendix A

Objections to Expanding Voluntary E-Verify and Employment Verification in General

Objection: It is inefficient to implement employment verification for all employers; instead, verifi­cation should be limited to industries where the hiring of illegal immigrants is most prevalent.

According to this argument, the employment of illegal immigrants is concentrated in a very small num­ber of industries and can be prevented by instituting employment verification in those industries alone. Eventually extending verification to all employers would create substantial burdens on large numbers of legitimate employers and employees with very little added benefit. This argument is faulty for four reasons.

First, while illegal immigrants are disproportionately low-skilled, they are not, despite conventional wisdom, working in only a handful of industries and jobs. For example, illegal immigrants work most commonly in construction, restaurants, landscaping, janitorial services, food processing, and hotels, but only around half of illegals work in these fields; the rest are spread through the economy. Some 45 per­cent of illegal aliens are visa overstayers rather than unlawful border crossers, and 10 percent to 15 per­cent of all illegals actually have a college degree. Illegal immigrants are thus spread more widely through the economy than popular wisdom suggests.

Moreover, employment verification through E-Verify is effective only against on-the-books employ­ment. Such hiring is more likely to occur in higher-wage and higher-skill jobs and less likely to occur in low-wage fields such as construction and farm labor. This means that on-the-books employment of illegal immigrants is less likely to be concentrated in a few low-wage fields than is off-the-books employment.

For these reasons, employment verification that is limited to only a few industries is likely to miss most illegal employment and unlikely to serve as a substantial deterrent to illegal immigration. By contrast, employment verification across all industries, coupled with procedures to limit identity fraud, is likely to affect at least half of all illegal workers in the U.S.

Second, employers in industries such as construction and agriculture will object if employment verifica­tion is imposed on their industries alone. They will argue that such selective enforcement is discrimina­tory. By contrast, business has shown a greater willingness to accept employment verification if it is required uniformly for all employers.[47]

Third, current law states that it is unlawful to employ illegal immigrants in the U.S. in its entirety. It does not state that it is unlawful to hire illegals only in construction, agriculture, and hotel work but law­ful elsewhere.

To create a policy that implicitly ignores or condones the hiring of illegals in most industries but is very tough on such employment in a few fields undermines respect for the law itself. By conveying the implicit message that the government was not really serious about enforcing the bar against hiring illegals, such a policy of highly selective enforcement would invite the public to regard the law cavalierly, thereby eroding support for the entire strategy. For example, it would be impractical to enforce stiff penalties on the hiring of illegals in a few industries while completely ignoring such hiring in other fields. Effectively limiting the hiring of illegals can occur only if society at large respects the law and takes the policy seri­ously. This will not occur if the law is deliberately not enforced in industries where millions of illegals actively work.

Fourth, the failure to enforce the current law is unfair to lawful immigrants, citizens, and law-abiding employees. It rewards those who break the law and penalizes those who abide by the law. In the current situation, for example, an individual with a temporary visa who remains in the U.S. unlawfully after his visa has expired will be rewarded by the higher income that comes with his presence in the U.S., but a similar individual who respects U.S. law and leaves the country when his visa expires will be punished by a loss of income because he obeyed the law.

This is wrong and unfair. We should not have a de facto policy that creates financial rewards for those who break the law and penalties for those who obey it. The unfairness becomes even more extreme if the lawbreaker is granted amnesty and permanent citizenship while the immigrant who lawfully left when his visa expired may be unlikely ever to become a citizen.

Similarly, a policy of accepting employment of illegals in a given industry rewards law-breaking employers in that industry by giving them a competitive advantage and punishes the majority of employ­ers who obey the law. The U.S. should have a policy of universal fairness, not selective fairness. It should encourage broad respect, not general disregard, for the rule of law.

Objection: Broad-based use of E-Verify is not necessary to stop the employment of illegal immi­grants; use of existing Social Security mismatch information will be sufficient.

Each year, SSA issues "mismatch" letters to employees who have identifying information on their W-2 tax forms that does not match SSA Numident files. Some letters are also sent to employers with large numbers of mismatched employees.[48] Some have argued that the government should use these mismatch data rather than E-Verify to prevent employment of illegals, but such a system would not solve the prob­lem because the mismatch data are taken from W-2 tax forms, which are received at the end of the year. Under such a system, unauthorized workers could be employed for a year or longer before any action to stop the employment was initiated.

Objection: E-Verify creates privacy concerns because it allows employers (or individuals posing as employers) to access employment records or other information from the Social Security Administration.

E-Verify does not allow employers to examine information from the Social Security Administration or other government agencies. It simply allows the employer to corroborate that the name, date of birth, and Social Security number submitted by an employee matches a record in the SSA or USCIS files.

It would be impossible to mine or prospect for identity information through E-Verify. Employers or other individuals cannot use the system to obtain private information about individuals against their will. For example, if an employer knew someone's name and date of birth and wanted to obtain that individ­ual's Social Security number for illicit purposes, the employer could not obtain that number from E-Ver­ify; to do so, the employer would literally have to submit millions of random Social Security numbers for the individual to E-Verify until one of them accidentally received a positive confirmation. SSA and DHS could immediately detect that sort of abuse.

Objection: Americans should not be required to prove to the government that they are eligible for work in order to obtain a job.

In fact, for over 20 years, all Americans have been required to provide evidence of their eligibility for employment. During that time, employees have been required to show their driver's licenses or other doc­uments to demonstrate that they were legally eligible for work. The employer copies information from the documents provided by the employee onto the official I-9 forms, which then collect dust in the employer's file cabinets.

This process has not worked, because the government has lacked the means to examine and verify the accuracy of the information on the I-9 forms. E-Verify is a promising solution. It simply translates the information on the I-9 into an electronic format, thereby allowing the government to determine the accu­racy of the information quickly and efficiently and then act on that determination.

Some argue that E-Verify would create an unacceptable barrier between U.S. citizens and jobs. All Americans have the right to seek work and (according to this argument) should not be required to prove to the government that they are eligible for work. But they already do just that.

While proponents of this view may agree that it is proper for the government to prohibit the employ­ment of illegal aliens, they feel that it is improper for the government to check the authenticity of claims of citizenship in order to carry out this prohibition. That argument is illogical: It is impossible to prohibit the employment of illegals without determining who is a citizen or lawful resident and who is not.

To obtain employment, most illegals submit documents indicating that they are either citizens or law­ful immigrants. In order to prohibit the employment of illegals, the government must be able to separate those who are truly eligible to work from those who are not. To accomplish this, the government must be able to verify the information provided to the employer by the prospective employee.

Objection: Requiring E-Verify coverage for all employees would entail enormous government costs.

Requiring E-Verify checks for all new hires and all existing employees would entail around $1 billion in new government costs over a four-year implementation period. This amounts to about $2.50 per U.S. household per year.

Objection: Mandating universal E-Verify coverage for all employees would impose inordinate costs on business and would be especially burdensome to small business.

The market price for checking individuals with E-Verify is around $10 per employee. Business hires around 63 million new employees per year. If all new hires were checked by E-Verify, even at a higher cost of $20 per employee, the total annual cost to business would be around $1.2 billion per year. This amounts to about $12 for each household in the U.S. It is always important to be cautious about adding new regulatory burdens and/or costs to business.

The estimated costs of implementing E-Verify appear to be modest in comparison to the magnitude of the illegal immigration problem. Moreover, E-Verify and the current new-hire data collection system for child support collect very similar information from business. If the existing new-hire data collection were incorporated into a universal E-Verify system, the added costs to business to operate E-Verify could be significantly reduced.

Objection: Errors in the E-Verify databases will cause many citizens and lawful immigrants to be barred from employment.

In recent years, E-Verify has delivered an erroneous tentative non-confirmation to around one in every 250 U.S. citizens or lawful work-eligible immigrants screened by the system. These erroneous tentative non-confirmations required the individual to correct errors in his Social Security files. Correcting these errors almost always involves a trip to a local SSA office. Recent modifications in the E-Verify process should cut the erroneous response rate at least in half.

Additionally, an estimated one in 300 U.S. citizens or lawful work-eligible immigrants receives an erroneous final non-confirmation from E-Verify because he did not contest an earlier tentative non-confirmation. This problem can and should be reduced by providing individuals with a greater opportunity to review and correct their personal Social Security information files.

Objection: E-Verify is ineffective because it cannot prevent employment by identity fraud.

Identity fraud will never be completely eliminated. However, the supplemental procedures advocated earlier in this paper could be highly effective in blocking employment of identity thieves.

Objection: Expanding E-Verify could cause the system to crash because of inherent technological problems of scalability.

Not true. E-Verify currently processes around 5 million inquiries per year. If the government mandated coverage of all new hires, the result would be over 60 million inquiries per year. Some have expressed concern that this scale of expansion might cause problems with the underlying computer hardware; how­ever, USCIS has tested the E-Verify system's ability to expand. Those tests showed that, with the addition of five servers, E-Verify could process up to 240 million inquiries per year.[49]

Objection: E-Verify creates privacy problems by enabling the government to collect new informa­tion on people and permits DHS to access SSA data.

Not true. The only personal information entered into E-Verify is the employee's name, date of birth, Social Security number, and citizenship status. This information is already included in the official I-9 forms that the employer completes for each employee. E-Verify merely allows USCIS to check the infor­mation on the I-9 in a secure, encrypted, efficient, and timely manner. Moreover, the government already has this information in its records and routinely collects similar information as part of the new-hire pro­cess and for purposes of collecting income and FICA taxes.

Less than 200 DHS employees (in USCIS) operate the E-Verify system in cooperation with SSA. These employees have access only to the information submitted through E-Verify and the SSA confirmation or non-confirmation of that information. They do not have access to the larger SSA employment history and earnings files for individuals. DHS employees cannot view or examine SSA records; they can merely cor­roborate that the limited identity data submitted for an individual through E-Verify do or do not match information in the SSA files.

The Westat evaluation of E-Verify concluded that the system created "little increased risk of misuse of…information by federal employees."[50] In addition, "it is unlikely that the program increases the likeli­hood of misuse of personal information."[51]

The SSA Numident data files represent the only means by which DHS can reasonably determine whether an individual is a citizen. Preventing DHS from corroborating information against SSA records in an efficient manner to determine the legitimacy of citizenship claims would scuttle any meaningful effort to limit the employment of illegal immigrants. Specifying in law that DHS personnel can use SSA data only for definite enforcement purposes might assuage privacy concerns, although this principle is already implicit in E-Verify operations.

Objection: Federal and state governments already collect information on newly hired employees for the National Directory of New Hires (NDNH) database, which is used for child support enforce­ment; the new hire data for the NDNH, not E-Verify, should be used for employment verification.

The new-hire reporting system is maintained as part of the National Directory of New Hires by the Office of Child Support Enforcement (OCSE) at the U.S. Department of Health and Human Services. This is a slow, cumbersome, and complex system that cannot realistically be used for employment verification; nor is it designed to do so.

Under the current new-hire data reporting system, employers provide information on recently hired employees to state government agencies. (In some cases, the agency may be a private contractor hired by the state.) The state agency then sends the employee data to OCSE in Washington. OCSE sends the infor­mation to SSA to check for Social Security mismatches. SSA returns the checked data to OCSE, which sends the information back to the state agency, which in turn communicates with the employer.

The time lag for the initial submission of data alone can be 32 days or longer. (Employers have up to 20 days after hiring to send data to the state agency, which has eight days to submit the data to OCSE, which in turn has four days to enter the data into the NDNH database.)[52] The time frame for communi­cating mismatch data back to employers is unclear. Employers can submit information by a variety of means; by law, they have the right to send all information in written form by physical mail to their state agencies.

Providing mismatch data is incidental to the main function of the new-hire data system; many states do not bother to communicate this information to employers at all. The new-hire process is too slow and error-prone to be used for employment verification, and the idea of interposing unrelated state child sup­port agencies (or independent contractors hired by the states) as unnecessary intermediaries channeling verification communications between SSA and DHS and employers is both inefficient and unworkable.[53]

Still, the E-Verify and new-hires reporting systems do collect similar information. If electronic employ­ment verification became mandatory for employers, E-Verify and new-hires data collection could be inte­grated, thereby eliminating the need for employers to submit the same data twice; however, this integration should be accomplished by upgrading and streamlining the new-hire data collection through E-Verify rather than saddling DHS with an antiquated, cumbersome data process that is entirely unsuit­able for verification purposes.[54]

This means that any new law requiring employers to use E-Verify should also eliminate the present fed­eral requirement for states to collect new-hire data for child support. Instead, all new-hire data should be collected directly by E-Verify and transmitted (along with information on final non-confirmations and Social Security mismatches) from SSA and DHS to the NDNH at OCSE. OCSE would then share the information with state child support agencies. The need for states to collect new-hire data independently and send them to SSA for corroboration would be eliminated.[55]

This system would meet the needs of both employment verification and child support enforcement. By contrast, creating an employment verification system that used antiquated new-hire technology and inter­posed state child support offices as unnecessary intermediaries between SSA and DHS and employers would be a recipe for failure. Replacing E-Verify with the new-hire data collection system would be like replacing a stealth bomber with a Model T Ford.

Appendix B

Accuracy of Numident Data

The Office of the Inspector General of the Social Security Administration divided its audit into three groups: native-born citizens, foreign-born naturalized citizens, and lawful residents.[56]

Native-Born Citizens

These individuals comprise 87.5 percent of the Numident files. While the OIG audit found data dis­crepancies in 3.3 percent of the Numident files for native-born citizens, nearly all of these were potential false positive discrepancies. Only 0.5 percent of the Numident files of native-born citizens contained false negative errors that would result in the individual being issued a temporary tentative non-confirmation from E-Verify.

Typically, these false negative errors involved name misspellings or clerical errors concerning date of birth in the Numident file or a failure to update a woman's name after she had changed her name through marriage. All of these errors were the sort that could easily be corrected by a short visit to the local Social Security office, thereby eliminating the tentative non-confirmation from E-Verify. Most of these errors would eventually need to be corrected by the individual under any circumstance.

Another 3.1 percent of Numident files of native-born citizens contained potential false positive errors. All of these errors involved individuals who appeared to be deceased but were not recorded as dead in the Numident database; many had a "tentative" but not a final deceased entry in their files.

The fact that large numbers of deceased persons are registered as potentially alive in the Numident files would seem to be a large loophole through which illegal aliens could falsely be verified for employment; however, exploiting that loophole would be difficult. An illegal immigrant would need to know the name, date of birth, and Social Security number of the deceased person as well as the fact that his demise was not recorded by Numident. The OIG report suggested that this potential loophole could be further nar­rowed by having E-Verify issue a tentative non-confirmation for individuals who had tentative death notices in their Numident files.

Naturalized Citizens

Around 1.8 percent of the Numident files were individuals recorded as naturalized citizens. OIG found discrepancies in 3.1 percent of these files "that could result in incorrect employment eligibility feedback" through E-Verify. OIG also found that 0.7 percent of the files of naturalized citizens contained clerical errors such as incorrect date of birth that would result in a false negative tentative non-confirmation from E-Verify.

OIG did find that 2.3 percent of individuals listed as naturalized citizens in Numident were not citi­zens but immigrants whose work eligibility was unclear. E-Verify would issue a false positive confirma­tion in these cases. Although these errors do represent a vulnerability in the operation of E-Verify, it is a small one. If E-Verify were used for all 63 million new hires in the U.S. each year, only about 24,000 unlawful workers could be expected to slip through this loophole.

Finally, the OIG found that 5.3 percent of the "naturalized citizens" in the Numident files were not cit­izens but legal permanent residents or other immigrants authorized to work in the U.S. E-Verify would give a correct positive confirmation authorizing employment to these individuals even though the under­lying citizenship data were wrong. Many of the errors concerning citizenship status involved older data entered before Social Security began to corroborate immigration data with the Department of Homeland Security.

Non-Citizens

OIG found that non-citizens comprised 10.7 of all Numident cases. The OIG audit found discrepan­cies in 10.4 percent of these records. Two-thirds of these discrepancies involved individuals who had been naturalized but were still shown as legal permanent residents in the Numident files. Changes imple­mented in May 2008 should reduce the number of tentative non-confirmations issued by E-Verify for these cases.

The remaining one-third of discrepancies (involving about 3 percent of all non-citizen files) involved cler­ical errors such as misspellings of names or errors in date of birth. Although E-Verify would issue an errone­ous non-confirmation for employment in these cases, this situation could easily be resolved by a brief visit to a local Social Security office. This small remaining problem could be greatly reduced if individuals were provided with an opportunity to check their Numident data independent of the hiring process.