Order online or call 1-800-767-1241

CJ Section: 
Author Name(s): 
Jim Harper

Jim Harper*

Successful "internal enforcement" of immigration law requires having a national identity system. If expanded, "E-Verify," the much-debated effort to control illegal immigration through access to employment, will become such a system, and it could easily be converted to controlling many dimensions of Americans' lives from Washington, D.C.

The neat congruence between expanded internal enforcement of immigration law and the national identification system necessary to implement it raises new questions about the advisability of worker background checks. A better approach than scouring the economy and society for people who illegally traversed the border might be substantive immigration policy reforms that are more consistent with liberty and prosperity for all.

What Is a National ID?

In past decades, gut feel and consensus have been decent barometers in debates about national identification, but it is worth defining the concept with at least some formality. Identifying people is a common and useful economic and social function, but there is a point at which the balance of power in an identity system shifts, and it no longer serves the individual's needs but binds the individual to the system by controlling access to economic, social, and political life. This is when an identity system works counter to individual liberty. It may be possible for a private identity system to have these characteristics, but historically and in all but the least likely future scenarios it is the coercive power of government that produces such an identity system.

The definition of "national ID" I have developed over time has three parts (Harper 2009): First, it is national. That is, it is used throughout the country and nationally uniform in the identifiers it uses and their formats. Second, its use is either practically or legally required. Things like credit cards are national, and some can be used to identify, but people can rotate among credit card systems or navigate the society without having any such card. The final part of a national ID is its use for identification. This is distinct from a national identifier such as the Social Security number, which merely attaches a number to a name. An identification system creates a high level of confidence that a particular individual, previously known to the system, has presented him-or herself once again (Harper 2006).

A system that is nationally uniform, practically or legally required, and identifying is a national identification system. In examiningE-Verify and the policy trajectories of internal enforcement, the job is to watch for these three elements coming together.

The American polity has long resisted a national ID, but federal immigration policy has been trending in that direction for a couple of decades. In the 1980s, attorney general William French Smith advocated in a cabinet meeting that a national ID card should be instituted for illegal immigration control. President Ronald Reagan reportedly scoffed, "Maybe we should just brand all the babies" (Derbyshire 2001). But in 1986 immigration reform law took a modest step in the direction of a national ID with the debut of "internal enforcement" of immigration law.

A Short History of Internal Enforcement

Though many take it as a given, internal enforcement of immigration law through worker background checks is an experiment that is only a few decades old. So far, it has failed to yield much in the way of results.

The Immigration Reform and Control Act of 1986 changed the long-standing, natural rule that working in the United States depended simply on willingness and ability plus an employer willing to hire. IRCA made unlawful the knowing employment of workers who are not eligible to work in the United States under the immigration laws, and it required employers to perform cursory checks of workers' documents upon hiring them. All employers today are required to verify employees' work eligibility by collecting completed I-9 forms and by checking employees' documentation.

Ten years later, with illegal immigration continuing apace, the Illegal Immigration Reform and Immigrant Responsibility Act of 1996 sought to "improve on" the policy of internal enforcement. It required the Immigration and Naturalization Service to commence three pilot programs to test electronic verification of employees' work eligibility. Among them was the Basic Pilot Program. Basic Pilot - recently renamed the "employment eligibility verification" program, or EEV, and then renamed again, "E-Verify" - is the remaining effort to verify work eligibility electronically (Harper 2008).

E-Verify works as follows: After collecting I-9 forms, participating employers enter the information supplied by workers into a government website. The system compares these data with information held in Social Security Administration and Department of Homeland Security databases. If the name and Social Security number pairs match to citizen data at the SSA, a worker is approved. The system compares information from noncitizens with DHS data to determine whether the employee is a naturalized citizen or immigrant eligible to work.

E-Verify electronically notifies employers whether their employees' work authorization is confirmed. Submissions that the automated check cannot confirm are referred to U.S. Citizenship and Immigration Service staff in the DHS, who take further steps to verify eligibility or to find the worker ineligible.

When E-Verify cannot confirm a worker's eligibility, it issues the employer a "tentative nonconfirmation." The employer must notify the affected worker of the finding, and the worker has the right to contest his or her tentative nonconfirmation within eight working days by contacting the SSA or DHS. When a worker does not contest his or her tentative nonconfirmation within the allotted time, the E-Verify program issues a final nonconfirmation for the worker and the employer is required to either immediately terminate the worker or notify DHS that it continues to employ the worker - confessing to a law violation.

The logic behind "internal enforcement" is simple: making it illegal to hire an illegal immigrant can reduce the strength of this country's economic "magnet," and fewer workers will try to come to the United States. But the policy built on this simple logic has generally failed. Just as a magnet's attraction passes through paper, the attraction of the United States to immigrants surpasses this paperwork.

The I-9 process and employer sanctions undoubtedly had some effect on illegal immigration and working, but not very much. Between 1986 and 1996, illegal immigration rates appear to have remained steady. Document fraud undermined the I-9 system, and the law prompted some employers to discriminate wrongly against citizens and legal immigrants because of their Hispanic surnames, poor English-language skills, or appearance. E-Verify has strengthened this system in some respects, but has not changed its fundamental character in treating work as an entitlement administered by government rather than the essential right to earn one's living.

How to Administer a Federal Personal Entitlement

Rights are freely exercised unconstrained by government, but entitlements must be administered and meted out as public policy requires. As noted above, the Immigration Reform and Control Act converted the right to earn a living into a federal entitlement that turned on immigration status. So how would a government-granted entitlement to work be administered?

To be more precise, IRCA made working legally in the United States contingent on presenting a certain credential: proof of legal eligibility. There are two steps to administering that credential: 1) identifying the individual proposing to work and 2) determining that individual's eligibility. The first step - identification - is surprisingly fraught when workers enjoy a substantial reward for defeating identity checks, and when employers have little interest in preventing identity frauds.

Our success and familiarity with in-person, customary identification gives many people excessive confidence in identification's power. In personal interactions, people use identification constantly. We are very adept at recognizing others with our eyes, ears, and other senses. That skill enables us to pick up right where we left off when we see people a second, third, and fourth time (Harper 2006).

The employment relationship is not like relationships with friends and neighbors, though. It typically commences among strangers, and in low-skill jobs, particularly, the new employee proffers his or her identity for the first time at the beginning of the relationship, often to a human resources manager with whom he or she will rarely interact. The employer takes little time to examine the applicant's identity bona fides.

At this early point in the relationship, though, the government requires the employer to examine and report on the new employee's identity information. It is not a natural, personal interaction of the kind that gives us so much confidence in identification. It is identification by card.

In my book, Identity Crisis: How Identification Is Overused and Misunderstood (Harper 2006), I described the process of identifying someone by card. This is an important and valuable economic and social function, which allows people to be treated as "known," at least to a degree, from the first encounter. But the three-step identification-by-card process is also full of weaknesses. First, the subject applies to a card issuer (such as a Department of Motor Vehicles) for a card, typically supplying all the information on the card. Next, the card issuer creates a card, supplying information to any later verifier. Finally, at the point of actual identification, the verifier compares the card to the subject and, having verified that the biometric information on the card is about the subject, accepts the proffered identity and other information on the card (Harper 2006).

Each of these three steps is prone to error or fraud, an opportunity for false negatives to creep into a system that searches for the work-ineligible. In the first step, the subject may apply to the card issuer with false information (including false documents), or the subject may corrupt employees within the card issuer, causing them to issue an inaccurate but genuine card. This will almost certainly deceive the employer, who cannot be expected to second-guess information printed on a genuine, un-tampered-with card.

At issue in the second step is the security of the card against forgery or tampering. Though many government-issued ID documents are quite resistant to forgery and tampering, the broadened use of these documents (including for immigration control) has increased the value of forging such documents and devising ways to tamper with them. Employers, who would be acting against their own interests to discover such things, cannot be expected to discover forgery or tampering of any decent quality.

The third step, comparing the identifiers on a card to the subject, is an area where employers are not specially disabled - everyone has the same ability to compare a picture to the face in front of them - but it is surprisingly difficult to compare a small photo on a card to a person similar in appearance, especially with changing hairstyles, variable lighting conditions, and the inexperience that many majority-race individuals have with the facial features of minorities. Again, employers will not be terribly eager to discover deception. A wrongful accusation would be extremely rude if incorrect, and - worse - it would invite a lawsuit.

Considering the many tenuous steps involved in the process, it is little wonder that deputizing employers as immigration agents has not been an end to the U.S. economic "magnet." There are high false negatives in the system today, allowing people not entitled to work under the law to labor productively all the same. Thus policymakers press for improvements.

"Improving" E-Verify: Toward Biometrics and a Cardless National ID

Given the flaws in the current system, it is unsurprising that there should be a push to improve it. However, improvements designed to strengthen the identification checks in E-Verify essentially require having a national identification system. Almost no way exists to do national employment eligibility checks that is not a step down that road. Taking the steps in identification-by-card in reverse order, we can see the direction the internal enforcement policy leads: machine biometrics1 and ultimately a cardless national ID.

The last step in the identification-by-card process, again, is the verifier's comparison between the biometrics on the card and the person presenting it. Rather than relying on the fallible human perception used in verifying photo-ID cards, an "improved" E-Verify system might use modern biometric measuring of each American, such as by fingerprint readers or iris scanners. Were every American biometrically registered, the biometric information embedded in their cards - retrieved from the card, retrieved from the human, and compared by machine - could provide much stronger assurance to verifiers that cards are about the people presenting them. The U.S. federal government has many projects under way to advance the use of biometrics, including the Department of Homeland Security's Biometrics Working Group and a biometrics clearinghouse at Biometrics.gov. Legislation House Judiciary Committee chairman Lamar Smith (R-Tex.) introduced in the 112th Congress called for the creation of a "biometric employment eligibility pilot program" to amend E-Verify while expanding its use (U.S. Congress 2011).

The security of cards against forgery and tampering (the concern raised in the second ID-by-card step) can be improved vastly with a variety of techniques, especially by using encryption. Printing or embedding information in a card using cryptographic techniques can establish with a high degree of certainty that the information was placed there by a particular agency and that it has not been altered since it was placed there.

Such security techniques do not have to incorporate identifiers, though they often do. Doing so can conceal from the ID subjects themselves what is on the cards they carry and display. This reduces the control the individual has over the process in a subtle but important sense. A person may share information in a powerful digital format that he or she might prefer not to share or prefer to share only in an obscure analog format. The REAL ID card standard, for example, provides for race information to be stored on the card and communicated digitally (Harper 2007).

But it is the first step in identification-by-card where the most "improvement" is needed - and where the national ID implications are most profound. Recall that this step requires ensuring the veracity of the information that makes it onto a card. This is challenging. After all, along with the biometrics, the information on identity cards is a collection of biographic data - name, date of birth, address, and such - that is not easily verifiable. Names themselves are pulled from the air by parents. They have no substance of their own or fixity to the body. They change or hyphenate as people marry, divorce, and remarry, for example, or by court decree. People move.

The way to ensure that accurate data are found on a card is to abandon the current practice of allowing applicants for identity cards to submit information about themselves. Rather, identity information and records of important life events could be cataloged from birth (or a person's first entry into the country). This has begun as part of compliance work on the REAL ID Act, the United States' national ID law, by the National Association for Public Health Statistics and Information Systems (NAPHSIS). Its Electronic Verification of Vital Events (EVVE) system "allows immediate confirmation of the information on a birth certificate presented by an applicant to a government office anywhere in the nation irrespective of the place or date of issuance," according to the NAPHSIS EVVE website (NAPHSIS 2011). The organization boasts that its system can match 250 million birth records in vital record databases nationwide.

The EVVE system is also capable of supporting the electronic verification or electronic certification of death records, and it could be used to verify or deny any data submitted by applicants for identity cards, such as marriage or divorce, address and homeownership, Social Security number, prior addresses, and so on.

This system is still susceptible to fraud, of course. A person claiming a series of identifiers that match EVVE records has only proven that he or she has good information on the identity he or she claims. Compiling such data on a large scale is the next step in identity fraud.

The solution to growing sophistication among fraudsters is to use some form of biometrics. Applicants for cards could submit some of the information, but the core identity information on each card would have to be tied to a central biometric identity repository - probably run by the government or by a government contractor like NAPHSIS. Reliable biographic information would have to be collected and stored by this repository or on a network of trusted parties that it runs. This system would reduce fraudsters' ability to submit false information to ID issuers.

The E-Verify photo-screening tool is a step in this direction, beginning to make the digital dossier as important as the card. In this program, when a noncitizen new hire presents a DHS-issued permanent resident card or employment authorization document, employers compare the photograph on the card presented to a copy of the photograph appearing on the employer's computer screen via the E-Verify system. If the photographs do not match, the employee is issued a tentative nonconfirmation (USCIS 2007).

Thus we see biometric information migrating from the card to the government's databases, where it can be transmitted for identity checks anywhere, outside of the control of the individual. The Department of Homeland Security desires to collect passport photos from the State Department and driver's license photos from DMVs around the country to expand the photo screening tool from using DHS-issued documents to using all Americans' passports and driver's licenses. In May 2011, the DHS announced a plan to add records derived from state DMVs to the E-Verify system through a network operated by the American Association of Motor Vehicle Administrators (DHS 2011a). That plan does not include transfer of biometric data like driver's license pictures, but it is the path along which such data will flow when new challenges require new "improvements" to E-Verify. Rather than being asked to present any document, employers would look up new workers on a government system to see if they are permitted to work.

This represents an awesome transfer of power. Because it is a database system, the individual would be in no position to argue his or her case as he or she can in the familiar card environment. With a card, a person can say, "No, look. This is me. This is my ID card. This is my picture. Give me my prescription." With E-Verify as a cardless national ID, the answer will be, "Sorry, you're not in the system. You have to talk to DHS or Social Security."

E-Verify and Mission Creep

Would E-Verify be used for prescriptions? Expanding E-Verify would not just increase the number of employers using E-Verify and the number or workers reviewed by the system. It would change a variety of behaviors in a variety of ways. For example, more employers and workers would collude to get productive work done "under the table." E-Verify expansion would increase identity fraud and deepen it, taking identity fraud into the complex world of biometrics, for example, so that the rewards of defeating the "strengthened" system could be reaped.

The expansion of E-Verify would also change the behavior of governments. Were an electronic employment verification system in place, governments could easily extend it to other uses. Failing to reduce the "magnet" of work, electronic employment verification could be converted to housing control, for example. The government might require landlords and title companies to seek federal approval of leases and sales so as not to give shelter to illegal aliens. Electronic employment verification could create better federal control of financial services and health care, to name two more.

There is no reason why a system created for immigration control would not be converted to other purposes, of course. Electronic verification could be used to find wanted murderers by requiring E-Verify checks everywhere from bus depots to liquor stores. Nationally mandated E-Verify would move quickly down the scale of uses to enforcement of unpaid parking tickets and "use taxes," just to name two examples. Instead of the card-based identity checks used at airports, travelers might be run past E-Verify as they proceed to airport concourses. Their records could be scanned for outstanding violations and late tax payments so they could get square with the government before they are allowed to travel.

The list of ways E-Verify could be used is limited only by the imagination. Electronic employment verification charts a course for expanded federal surveillance and control of all Americans' lives.

Toward an Immigration System that Embraces Freedom

Until the 1986 immigration reform law, decisions about who should work for whom were made between employers and workers. Earning a living was treated as a right, and sensibly so given its centrality to sustaining one's own life and the lives of one's family members. Letting workers and employers get together on their own terms makes eminent sense, just like people deciding for themselves what food they should eat and how their kids should be schooled. But in the name of immigration control, Congress began the shift toward treating working as an entitlement controlled by government and administered by employers.

Even moving forward as a mostly voluntary program, E-Verify's implementation has been fitful and costly. In Arizona, a state that mandated E-Verify use in January 2008, only one-third of employers were submitting new hires to E-Verify by the midpoint of that year (Westat 2009: 87). The Social Security Administration itself failed to use the system for 19 percent of its new hires in fiscal years 2008 and 2009 (SSA 2010). Meanwhile, over 50 percent of unauthorized workers screened through E-Verify from April to June of 2008 were confirmed by the system though they were not supposed to be (Westat 2009: 51-53 and B10-B12).

E-Verify costs about $100 million per year currently (DHS 2011b: 9). Those costs would rise if the program expands. The Congressional Budget Office reported in 2007 that mandatory E-Verify would cost $3 billion over the 2008-12 period and $6.1 billion over 2008-17 (CBO 2007). Meanwhile, a national E-Verify mandate would drive down federal tax revenues by an estimated $17.3 billion over 10 years (Orszag 2008). The Migration Policy Institute, though it calls electronic eligibility verification "essential," says that "enactment of new E-Verify mandates without broader immigration reform may do more to harm the economy and U.S. workers than to deter illegal immigration and to protect U.S. jobs" (Rosenblum 2011).

These prohibitive costs may have a welcome aim. Whether stated exactly as such or not, the goal of many of E-Verify's proponents is to bring the rule of law to the immigration environment. Fealty to law is important for the maintenance of a just and stable society, and immigration law is widely disrespected and often broken. But good law is not a hammer waved over the heads of subservient people. Good law gives expression to the values of the people.

Immigration law is disrespected and broken not because it is poorly enforced, but because it is inconsistent with the will of the people. In the main, the majority of the American people express their will quietly but insistently in their decisions to hire good, hard workers, and to enjoy the product of these workers' labor, indifferent to where the worker was born.

Expanded E-Verify would produce a national ID system anathema to freedom. The policy of internal enforcement is necessitated by an immigration law regime that is also anathema to freedom. Once in place, use of such a system would metastasize into many uses that control American people's lives.

A better approach than E-Verify and internal enforcement would be substantive immigration reforms that produce policies more consistent with freedom of workers and employers to get together on the terms they want, including workers from nearby countries with the desire to work and the skills to lend to our economy. Rather than shrinking from freedom in order to shrink the economy using E-Verify, immigration reform should embrace freedom, including freedom of movement so that people from other lands can improve their lives and our economy by trading on their skills and labor.


Congressional Budget Office (2007) "Senate Amendment 1150 to S. 1348, the Comprehensive Immigration Reform Act of 2007." Washington: CBO (4 June). Available at www.cbo.gov/ftpdocs/81xx/doc8179/SA1150_June4.pdf.

Department of Homeland Security (2011a) "Privacy Act of 1974; Department of Homeland Security/United States Citizenship and Immigration Services - DHS/USCIS-011 E-Verify Program System of Records." Docket No. DHS-2011-0030), 76 Fed. Reg. 26,738. Washington: DHS (9 May). Available at www.gpo.gov/fdsys/pkg/FR-2011-05-09/html/2011-11291.htm.

___ (2011b) "Budget in Brief FY 2009." Washington: DHS. Available at http://www.dhs.gov/xlibrary/assets/budget_bib-fy2009.pdf.

Derbyshire, J. (2001) "Your Papers, Please." National Review (11 November). Available at www.johnderbyshire.com/Opinions/NationalQuestion/papersplease.html.

Harper, J. (2006) Identity Crisis: How Identification Is Overused and Misunderstood. Washington: Cato Institute.

___ (2007) "Understanding the Realities of REAL ID: A Review of Efforts to Secure Drivers' Licenses and Identification Cards." Testimony before the Senate Committee on Homeland Security and Governmental Affairs Subcommittee on Oversight of Government Management, the Federal Workforce, and the District of Columbia (26 March). Available at http://hsgac.senate.gov/public/_files/TestimonyHarper.pdf.

___ (2008) "Electronic Employment Eligibility Verification: Franz Kafka's Solution to Illegal Immigration." Cato Policy Analysis No. 612. Washington: Cato Institute (March). Available at www.cato.org/pub_display.php?pub_id=9256.

___ (2009) "Is the REAL ID Revival Bill, 'PASS ID,' a National ID?" Cato@Liberty (www.cato-at-liberty.org/is-the-real-id-revival-bill-pass-id-a-national-id).

National Association for Public Health Statistics and Information Systems (2011) "Electronic Verification of Vital Events (EVVE)." Silver Spring, Md.: NAPHSIS. Available at www.naphsis.org/index.asp?bid=979.

Orszag, P. (2008) Letter to Rep. John Conyers Jr., Chairman of the U.S. House Committee on the Judiciary. Washington: CBO (4 April). Available at www.cbo.gov/ftpdocs/91xx/doc9100/hr4088ltr.pdf.

Rosenblum, M. (2011) "E-Verify: Strengths, Weaknesses, and Proposals for Reform." Washington: Migration Policy Institute. Available at www.migrationpolicy.org/pubs/E-Verify-Insight.pdf.

Social Security Administration (2010) "The Social Security Administration's Implementation of the E-Verify Program for New Hires." Audit Report 2010, Office of the Inspector General. Washington: SSA. Available at http://oig.ssa.gov/social-security-administrations-implementation-e-verify-program-new-hires.

U.S. Citizenship and Immigration Services (2007) "USCIS Launches Photo Screening Tool for E-Verify Program." Washington: USCIS (25 September). Available at www.uscis.gov/files/pressrelease/EVerifyRelease25Sep07.pdf.

U.S. Congress (2011) "Legal Workforce Act." H.R. 2164 and H.R. 2885, 112th Congress.

Westat (2009) "Findings of the E-Verify Program Evaluation." Report submitted to U.S. Department of Homeland Security. Washington (December). Available at www.uscis.gov/USCIS/E-Verify/E-Verify/Final%20E-Verify%20Report%2012-16-09_2.pdf.


1"Biometrics" literally refers to the measuring of living things and so includes both analog and digital techniques, but in common parlance analog identification techniques such as human visual assessment are not called "biometrics" and digital techniques are called "biometrics." They are all biometrics, but because of their greater scalability and consequences for data collection, retention, and processing, machine biometrics are more concerning.