November 2005 No. 6

SUMMARY

Gaining control of illegal immigration to the United States would require Immigration large-scale reforms to the notoriously ineffective employer sanctions system. Enforcement at Such changes should make it easier for employers and migrants to comply the Worksite: with the law, and more punitive for those who do not. Making it Work Marc R. Rosenblum A number of reform strategies have been proposed, each with its own I. Introduction economic and privacy costs. These include consolidating identification Most unauthorized immigrants enter the United States in pur- documents and making them more suit of employment, and analysts view an effective employer secure; mandating application screen- sanctions regime as an essential element of immigration con- ing through government databases; trol. Employer sanctions are also attractive to some advocates devoting additional staff to worksite of more open policies as part of a “grand bargain,” which enforcement; increasing penalties for combines tougher enforcement with more legal flows. Yet US noncompliance; and streamlining and employer sanctions laws have been notoriously ineffective strengthening the enforcement because employers lack tools to verify applicants’ eligibility, investigators lack tools to identify non-compliant employers, process to increase investigators’ and (for these and other reasons) employers view the threat of worksite access. actual punishment under the current system as an acceptable business expense. Two additional possibilities include creating a centralized screening sys- At least six different types of reform could strengthen the tem and a job holder database. These employer sanctions system: reforms could reduce the burden on • Improvements to document security, employers and the likelihood of dis- • Document consolidation, crimination against eligible employees • Mandatory use of employment databases, by “front loading” verification at the • Increased enforcement staffing, point of document issuance, and could • A revised penalty structure, and simplify eligibility verification and • Better worksite access for investigators. streamline enforcement. Such reforms, and their trade-offs, are essential con- Creating an effective employer sanctions regime would likely siderations for improving the integrity require some change within all of these areas, and bills under of the US immigration system. consideration by the 109th Congress encompass all but the last of them. Yet within each of these areas, (IRCA) has been the I-9 form. The form enu- the potential benefits of more effective merates eight documents establishing both enforcement are partially offset by economic identity and work eligibility, twelve documents and other costs of reform. Moreover, existing only establishing identity, and seven docu- congressional proposals may be overly opti- ments only establishing eligibility (current mistic about technological fixes to the prob- instructions on the USCIS website narrow the lems of document and identity fraud and list to twenty-five documents.) To demonstrate about the potential to implement a universal their legal status, employees are required version of the Basic Pilot employer verifica- either to produce one of the dual-purpose doc- tion program. uments (e.g., a US passport or Green Card) or one document from each of the other two lists Thus, in addition to reviewing the costs and (e.g., a state-issued driver’s license along with benefits of these incremental reform strategies, a Social Security card). IRCA requires this brief also argues for consideration of two employers to confirm—to the best of their additional kinds of changes to the existing ability—that new employees can satisfy this employer sanctions regime. First, rather than requirement, and employers document com- requiring employers to query an eligibility data- pliance by maintaining completed I-9 forms base at the point of hire, a centralized screen- on file for every employee. ing system should be established in which the verification process occurs at the point of docu- The optional Basic Pilot employment verifi- ment issuance. Second, employers’ primary cation program, established in 1996, com- responsibility would then be to submit employ- plements the I-9 process by allowing ees’ identification information to a job holder employers to compare a job applicant’s database. This database would become a pri- identification information to the Social mary tool of worksite enforcement, and would Security Administration’s (SSA) database also enhance the integrity of the United States’ and the Department of Homeland Security’s overall system of migration control. (DHS) immigrant database. Where neither II. Background:The Existing automatic nor manual searches of these Employer Sanctions Regime databases are able to confirm a worker’s eli- gibility, the Basic Pilot issues a final non- Any system of employer sanctions must begin confirmation of eligibility and employment with procedures to confirm the eligibility of must be terminated. prospective employees, a process that includes two distinct but related tasks: Following this verification process, sanctions enforcement consists of three tasks: • matching an individual with his or her name and relevant personal information • Targeting employers for investigation. (identification), and Most sanctions investigations have • confirming that the individual is authorized focused on individual businesses on the to work in the United States (verification). basis of leads provided by private citi- zens, the Department of Labor, and In the United States, the primary tool for iden- other local, state, and federal law tification and verification since passage of the enforcement agencies. A secondary 1986 Immigration Reform and Control Act strategy has been to target firms on the

2 Policy Brief

basis of their industrial and regional rized immigrants also borrow or steal legitimate characteristics. Some firms were ran- documents and obtain work under someone domly selected for inspection during the else’s name. These two problems—document late 1980s and the 90s, but random tar- fraud and identity fraud—are exponentially geting was discontinued in 1998 for lack more difficult to prevent as a result of I-9 com- of results. plexity: with at least twenty-five different docu- ments in the system, it is simply not possible • Investigating and prosecuting for employers who lack employers. Once a firm is targeted, inves- specialized training to With at least twenty-five tigations have primarily relied on audits of reliably screen job different documents in the I-9 forms and other records. Audits may be applicants. IRCA’s followed by worksite inspections to confirm system, it is simply not pos- well-intentioned anti- the correspondence between these records sible for employers who lack discrimination meas- and actual personnel. specialized training to reli- ures also undercut • Penalizing non-compliant employers. effective enforcement ably screen job applicants. Where investigators find evidence of by preventing careful non-compliance (historically slightly scrutiny of documents which appear to be gen- less than half the time), penalties may uine. Indeed, roughly half of all unauthorized range from a formal warning to a workers are hired by employers who fully com- “paperwork” fine (for unintentional non- ply with I-9 requirements. compliance), to a “substantive” fine (for intentional non-compliance), to the ini- The Basic Pilot program has not fixed these tiation of criminal charges (for engaging in problems. Most importantly, because the pro- a repeated pattern or practice of violations). gram is voluntary, it fails to detect the vast majority of ineligible applicants, including, in III. Flaws in the Existing System particular, those hired by intentionally non- compliant employers. The program is also Deportable aliens apprehended by the INS unable to detect the fraudulent use of bor- dropped by half in the three years following rowed or stolen documents, making it easily IRCA’s passage, and there is considerable evi- evaded by unauthorized migrants. dence that this reduction was partly a result of employer sanctions serving as a deterrent. Yet In addition, flaws in SSA and DHS databases by the early 1990s, employers and unautho- (along with employer errors) ensure that the rized immigrants learned to exploit IRCA’s Pilot produces a high level of false negatives, design flaws, especially with respect to eligi- or cases in which genuinely work-eligible indi- bility verification. IRCA enforcement has been viduals are not confirmed. These false nega- further undermined by statutes and regulations tives are particularly likely to affect recent that protect non-compliant employers. immigrants and other legal non-citizens. And with employers prioritizing rapid resolution The existing verification system rests on the over fairness, false negatives often lead to lost ability of employers to correctly screen employment opportunities. prospective workers for eligibility, but the widespread availability of counterfeit docu- Beyond the verification process, sanctions ments makes doing so difficult. Many unautho- enforcement is fundamentally undermined

3 by the fact that no agency, office, or division short-term disruptions in the availability of has made a priority of worksite enforce- immigrant labor, but failed to build employer ment. The Immigration and Naturalization support for sanctions or to achieve a lasting Service (INS) consistently placed greater deterrent effect.) A 1992 Memorandum of emphasis on the apprehension and removal Understanding sought to elicit tips from of unauthorized Department of Labor (DOL) agents, who Sanctions enforcement is immigrants than on make frequent worksite inspections, but fundamentally undermined worksite enforce- DOL commitment to protecting workers by the fact that no agency, ment. Indeed, the (regardless of their migration status) made it office, or division has priority placed on a reluctant partner. Thus, only 236 tips were worksite enforce- made a priority of generated by the collaboration between 1988 ment peaked during and 1998. worksite enforcement. the late 1980s at around 5 percent of Even when non-compliant employers are tar- the agency’s budget, and spending on inte- geted for enforcement, investigations have rior investigations increasingly lagged rarely produced convictions because of the behind spending on border enforcement and flawed verification system discussed above. In detentions during the 1990s following the short, as long as employers have completed an initial success of the Border Patrol’s “pre- I-9 form for every employee on the payroll, it vention through deterrence” strategy. is difficult for investigators to prove intention- Institutional commitment to employer sanc- al non-compliance. Investigations are further tions reached a new low after the 9/11 undermined by a statutory and regulatory attacks as the INS shifted its focus to terror framework designed to protect employers from prevention. Thus, less than 10 percent of over-zealous enforcement: immigration enforcement spending was ded- • 1996 Immigration and Nationality Act icated to interior investigations of any kind (INA) amendments give employers an in 2002; and only 2 percent of these interi- extra defense to a technical violation or or investigations targeted employers. The procedural violation if they made a shift in focus from employment to countert- “good faith attempt” to comply with errorism was formalized in 2003 when verification procedures; responsibility for interior investigations passed to the Immigration and Customs • 1998 regulations require warnings Enforcement (ICE) division within the (rather than workplace inspections) Department of Homeland Security. where investigators suspect that non- compliance was unintentional; and Beyond this general limitation, each aspect • 1998 and 2003 regulations require of enforcement confronts additional prob- inspectors to obtain detailed written lems. With respect to the targeting of approval from their supervisors prior to employers, reliance on tips has failed visiting worksites. because neither non-compliant employers nor unauthorized immigrants have incentives Finally, many non-compliant employers view to collaborate with enforcement agents. (INS sanctions as an acceptable business expense, efforts to foster employer collaboration reducing their deterrent effect. Fines range through industry-wide efforts like Operation from $100 to $1,000 per unauthorized immi- Vanguard in 1998-9 succeeded at creating grant for paperwork errors and from $250 to

4 Policy Brief

$10,000 for substantive violations—a statutory in Figure 1), targeting less than one-tenth of 1 range that has not changed since 1986 and percent of US worksites. Only 10 percent of which may be well below the cost savings from these cases led to final orders to fine (the bro- employing unauthorized labor. Moreover, many ken line), and an average of just $2.2 million migrant-employing businesses are small and in fines were collected (1991-1999). informal, and firms frequently rely on labor contractors to recruit their migrant workforce. IV. Potential Reforms As a result, collection of fines is impeded by relocation or restructuring of non-compliant Major migration proposals under consideration businesses, and INS/ICE lawyers often offer by the 109th Congress include a number of firms reduced penalties in an effort to increase measures that would strengthen the existing actual payment rates. employer sanctions regime. These proposed changes fall within six categories. Figure 1 illustrates the limits of employer sanc- tions enforcement. Between 1991 and 2003, an 1. Improve Document Security average of fewer than 5,000 employer investi- Identification and eligibility verification are gations was completed per year (the solid line undermined by document and identity fraud,

Figure 1: Employer Sanctions Enforcement Outcomes

Sources: INS Statistical Yearbook, various years; Peter Brownell, “The Declining Enforcement of Employer Sanctions,” Migration Information Source, Sept 1, 2005.

5 compromising the integrity of today’s employ- Technological solutions to the problem of er sanctions regime at every stage. Recent immigrants using borrowed or stolen docu- advances in document design hold out the ments are less promising, however, because promise of producing tamper-resistant, fraud- placing biometric scanning devices at a large proof identity and eligibility cards with bio- number of worksites would be prohibitively metric data linking them to their owners. In expensive, employers lack the necessary theory, secure I-9 documents could allow expertise to use biometric equipment, and no employers to become effective eligibility biometric database exists against which screeners and allow enforcement agents to prospective employees—i.e., every resident of hold non-compliant employers accountable. the United States—could be checked. Indeed, in light of the problems that the far less ambi- For these reasons, the four omnibus immigra- tious US-VISIT program has encountered as it tion proposals under consideration in the 109th seeks to confirm the identity of foreign Congress would seek to improve document nationals entering the United States, the security in the following ways: obstacles to biometric screening at worksites would appear insurmountable. More promis- • Add anti-fraud measures and biometric ing policy responses to this type of identity information to visas and immigration fraud involve better use of employment data- documents (Cornyn-Kyl and McCain- bases, discussed below. Kennedy proposals) • Add anti-fraud measures to Social What would be the costs of improving I-9 doc- Security Cards and improve security of ument security? Cost estimates for issuing birth certificates (Cornyn-Kyl and secure drivers’ licenses under the REAL ID Tancredo proposals) Act range from $125 million to $750 million • Increase penalties associated with for 200 million document fraud (Cornyn-Kyl and new licenses. With the passage of the REAL Tancredo proposals) These estimates ID Act, some move to enhance only cover the • Assemble a database on the production document security is now cost of anti-fraud and distribution of fraudulent documents inevitable, and the most measures, howev- (Jackson-Lee proposal) er, and fail to important questions regard the account for pro- scope of documents affected. Although the technology of “fraud-proof” cessing expenses. identity cards is far from perfect—well-funded A 1998 GAO study estimated the cost of issu- terror networks and other high-end consumers ing secure Social Security cards to all 270 would likely remain one step ahead of anti- million number-holders at roughly $4 billion, fraud efforts, for example—high-tech I-9 docu- with processing costs accounting for 90 per- ments could succeed at pricing phony docu- cent of the expense. This estimate does not ments beyond the reach of most unauthorized include the cost to individuals of lost work immigrants. Higher penalties on counterfeiters time associated with obtaining new cards. and anti-fraud law enforcement efforts are a natural and relatively low-cost compliment to Yet with the passage of the REAL ID Act, fraud-proof cards and should further reduce some move to enhance document security is their availability. now inevitable, and the most important ques-

6 Policy Brief

tions regard the scope of documents affected. fraud and an additional barrier to effective Enhanced security must extend to US citizens’ employer compliance. Document consolida- documents, especially given that recently tion and document enhancement are two sides issued green cards and immigrants’ work of the same coin. authorization cards already include biometrics and anti-fraud technology, while most citizens’ A One-Card System? Pursuing document documents do not. Yet in the interest of effi- consolidation to its logical end would lead to ciency and effectiveness, anti-fraud measures the creation of a new type of document: a uni- should not encompass twenty-five or more versal “work authorization” card. A one-card separate documents, but rather should be system would discard existing documents (for combined with steps to limit the number of purposes of eligibility verification) and imple- documents acceptable for verification. ment a uniform, fraud-resistant, machine- readable card to which all work- 2. Document Consolidation eligible individuals would be entitled. Cards The limits of anti-fraud measures point natu- would include prominent expiration dates rally to document consolidation. In short, the (marked “none” for US citizens and legal per- fewer documents involved in eligibility verifi- manent residents) reflecting the holder’s spe- cation, the less costly it would be to improve cific eligibility status. Cards could also their security, and the more feasible it would include biometric data, though most employ- be for employers to become reliable screeners. ers would be unable to take immediate advan- Current congressional proposals contemplate tage of this information. the following steps in this direction: • Restrict identification and verification doc- Moving to a one-card system has been resis- uments to enhanced Social Security cards ted because of widely held worries about cre- and a biometrically-equipped US govern- ating a “national ID card.” This phrase marks ment ID Card or REAL-ID compliant a widely voiced concern, and it has led some driver’s license (Cornyn-Kyl proposal) drafters of reform proposals to go to excessive • Restrict identification and verification doc- lengths to insure that employees can use sev- uments for non-citizens to enhanced visa eral cards to demonstrate work authorization. and immigration documents (McCain- In contrast to proposals which focus on visas Kennedy proposal) and migration documents, movement toward a national ID card would inherently affect all Numerous additional document consolidation Americans. While universal reforms of this schemes could be described. In general, with nature are clearly desirable from a migration every document struck from the list, the control standpoint, requiring citizens to share integrity of the system is enhanced and the the costs of document reform in this way burden of verification shifts from employers would be politically costly. (in the form of complexity) to workers (in the form of reduced flexibility). Thus, the Cornyn- Yet analysts should avoid allowing opposition Kyl proposal imposes an additional burden on to a “national ID card” overwhelm a discussion US citizens by prohibiting the use of a school of document consolidation, because ultimate ID as proof of identity, for example. The consolidation, if done with careful safeguards McCain-Kennedy bill would leave this privi- to minimize the privacy and civil liberties con- lege intact, but in so doing would also leave in cerns that underlie this opposition, would carry place an additional opportunity for document significant advantages. The most obvious

7 advantage of a one-card system would be its ly would increase to the extent that diverse simplicity: rather than being responsible for secure cards are bundled together, or to the reviewing twenty-five, fourteen, or two types of extent that a single card would come to be documents (including over fifty different vari- used for multiple ID and access purposes, as eties of driver’s licenses), employers would efficiency would suggest. But there are many only be required to familiarize themselves with possible variations, and it is certainly feasible one universal card. In this way, a one-card sys- to design a card that would be restricted to tem would make employer compliance far less certain specific and carefully chosen uses, and burdensome. Creating a new work authorization to combine a new card with enforcement card would also eliminate the logistical prob- efforts targeting misuse of identity data. lem of how to “grandfather out” existing docu- ments as their secure successors come on line. 3. Mandatory Electronic Screening What would be the costs and limitations of Currently, employers have the option of sub- moving to a one-card system? The primary mitting applicants’ identification data to the economic cost of a one-card system would be Basic Pilot verification system, but only a card issuance, and with processing costs small minority—one-twentieth of one per- would approximate the Social Security card cent—choose to do so. Increasing the pro- estimate of $4 billion. Some cost savings portion of employers who participate in such could be accomplished by consolidating the screening would strengthen the employer diverse secure documents under consideration sanctions system by improving the reliability (i.e., Social Security, driver’s license) into a of verification procedures and by creating single ID card usable for multiple purposes. new methods for targeting non-compliant employers. Yet a number of difficult obstacles Yet this observation raises a more fundamental to a universal eligibility screening system question: what is the broader cost of requiring must be overcome. individuals to possess a de facto national ID card in order to obtain employment? Many There is widespread support for a mandatory security experts believe such cards would pro- system of electronic verification, and bills in vide a false sense of security and divert atten- the 109th Congress would: tion from other law enforcement efforts while • Require mandatory employer participation providing few actual security benefits. (As in an electronic eligibility verification noted above, terrorists and professional crimi- system (Cornyn-Kyl, McCain-Kennedy, nals would almost certainly remain a step and Tancredo proposals; bills differ in ahead of anti-fraud technology.) details of how mandatory participation is phased in over time) Moreover, the existence of a national database • Require SSA (with DHS cooperation) to linking personal information to employment construct a new database specifically records (and possibly to driving and other designed for eligibility verification records) is an inherent threat to privacy. While (McCain-Kennedy proposal) security safeguards would be employed to min- • Require SSA to purge expired and incorrect imize the misuse of these data, outsiders could social security numbers from the NUMI- possibly gain access and exploit the informa- DENT database (Cornyn-Kyl proposal) tion in any number of ways. These privacy • Require SSA to search employer records costs are difficult to quantify, and they certain- for Social Security numbers which do not

8 Policy Brief

match employee names or are otherwise Reforming the Basic Pilot For these rea- incorrect and to identify Social Security sons, any move toward a universal electron- numbers associated with more than one ic verification system must be preceded by place of employment (Tancredo proposal) significant reform to the existing Basic Pilot. One alternative would be to retain the Simply put, if employers are to be charged existing two-database structure, but update with determining worker eligibility, they must SSA and DHS databases and procedures to be given—and required to use—the best reduce false negatives and speed response available tools for doing so. Mandatory elec- time. Yet such an effort has been underway tronic verification would allow compliant for five years, and holds limited additional employers to detect applicants’ use of coun- promise given the structure of the existing terfeit documents; and it would also prevent SSA database, which is not designed to other employers from going through the search for identity fraud, and does not motions of compliance to prevent prosecution. distinguish among migration categories. Yet requiring universal participation in the De facto discrimination will also persist as long existing Basic Pilot system would be costly as foreign-born applicants are more dependent given the high level of tentative non-confirma- on the DHS database for their confirmation. tions, which are now returned in a quarter of all cases (while less than 1 percent of queries Thus, the McCain-Kennedy proposal for a new result in final non-confirmations). Resolving eligibility database is a more promising, the status of these cases costs at least six dol- though more expensive, lars per case, and it would cost $10 million solution. A new data- The promise of data mining per year to conduct manual investigations in a base could be designed as a tool for targeting non- universal system at the current error rate. to accommodate regu- compliant employers depends More importantly, investigating non-confirma- lar updating of SSA on the quality and structure tions requires from twenty-four hours to two and immigration fields, of mandatory databases, but weeks, and these delays in a universal system and to search both sets would substantially slow the hiring process of data simultaneously. the ability of investigators to and reduce economic growth. Most importantly, a replace tip-based searches new database could be with “smart auditing” of In addition, as noted above, the Basic Pilot designed to detect non-compliant employers produces a high level of false negatives, and identity fraud by auto- represents an important these non-confirmations disproportionately matically flagging over- potential efficiency gain. affect foreign-born citizens and legal residents used identification whose status is unlikely to be confirmed by numbers for follow-up the more reliable SSA database. Moving to a investigation. Yet it should be emphasized that universal system would substantially exacer- any data entry errors, alternative spellings, bate this problem, creating a de facto system and delayed updates would continue to be of discrimination against the foreign-born. problematic in any verification system, and some mechanism would be required to Finally, the Basic Pilot is not designed to respond to the resulting errors. detect the use of borrowed or stolen docu- ments, so that even universal participation Electronic Verification as a Tool for would fail to reliably detect applications by Targeting Employers In addition to improv- unauthorized immigrants. ing the verification process, an improved

9 mandatory verification system could provide for employers to reliably verify applicants’ eli- an important tool for targeting employers. In gibility and the easier it is for enforcement short, just as data analysis should be used to agents to identify and successfully prosecute flag potential cases of identity fraud, so non-compliant employers, the fewer agents should analysts look for employers with suspi- would be required to deter illegal employment. cious hiring or query patterns (e.g., too many or too few queries, too many or too few tenta- The issue of adding agents raises the question tive non-confirmations, too many workers sus- of where responsibility for worksite enforce- pected of identity fraud, etc.). The McCain- ment should lie. Three main possibilities Kennedy bill contemplates this type of over- exist. First, sanctions enforcement could sight and would require DOL to monitor a new remain the responsibility of ICE, taking verification system in this manner. Ultimately, advantage of that division’s enforcement infra- the promise of data mining as a tool for target- structure and institutional culture. Yet few ing non-compliant employers depends on the unauthorized immigrants—including those at quality and structure of mandatory databases, critical infrastructure locations—represent a but the ability of investigators to replace tip- significant security threat, and ICE agents based searches with “smart auditing” of non- view worksite enforcement as a lower priority compliant employers represents an important than other, security-oriented efforts. Requiring potential efficiency gain. more worksite enforcement of ICE would divert resources from its security mission. 4. Increase Staffing Devoted to Employer Sanctions Second, primary responsibility for worksite Institutional inattention to employer sanc- enforcement could be shifted to the tions clearly undermines all aspects of sanc- Department of Labor’s Employment Standards tions enforcement, and two proposals in Administration. DOL is a logical choice for the 109th Congress would increase enforce- worksite enforcement because its existing ment resources: duties include tens of thousands of worksite • Add 2,000 ICE investigators annually for a inspections each year and because working period of five years (Cornyn-Kyl proposal) conditions would generally improve as unau- • Require a greater role for state and local thorized immigrants are removed from the law enforcement agents in immigration workforce. Yet sanctions enforcement also control in general (Cornyn-Kyl and conflicts with DOL’s mission of protecting Tancredo proposals) workers’ rights regardless of their immigration • Add 300 ICE attorneys over a period of status. The agency also resists a sanctions two years (Tancredo proposal) enforcement role for fear of discouraging workers from approaching DOL about other Each reform discussed in this brief would types of worksite violations. become more effective to the extent that it is backed up by additional enforcement person- A third option would be to create a new divi- nel. Yet deploying more agents without these sion, inter-agency taskforce, or some other additional reforms to strengthen the verification entity uniquely responsible for worksite and enforcement systems would be a waste of enforcement. Such an employer sanctions resources. Ultimately, the number of agents taskforce would avoid the problems of con- required to reduce unauthorized employment flicting institutional cultures—a cost associ- depends on these other features: the easier it is ated with DOL and ICE options that is diffi-

10 Policy Brief

cult to quantify. Yet building a new agency or be addressed by legislation to hold on-site taskforce from scratch would certainly be the employers and off-site contractors jointly most expensive option, and such a ground-up liable for employing unauthorized immigrants. operation would be unable to take advantage of existing infrastructure and economies of Yet imposing excessive minimum fines or scale within DOL or ICE. Thus, a new ICE increasing the likelihood of jail time may have division or unit with structural and budgetary the perverse effect of deterring investigators protections against diverting resources to and judges from imposing penalties. other investigations may be an attractive Employers prefer to avoid the inconvenience intermediate alternative. of the sanctions process, and regular enforce- ment of relatively low fines may be a more 5. Increase Penalties Associated with effective deterrent than sporadic enforcement Illegal Employment of higher ones. Another possibility would be to Fines for employing unauthorized immi- impose enhanced penalties on employers who grants—especially in the case of first offens- hire unauthorized immigrants while also vio- es—are almost certainly less than the cost lating other labor codes. savings to many employers of hiring unautho- rized workers. Bills in the 109th Congress The deterrent effect of stiffer penalties must would address this by increasing the level of be balanced against the problem of uninten- existing fines: tional non-compliance. As long as compliance is difficult, the presump- • Maintain a range of possible fines for A direct relationship tion of innocence requires paperwork, substantive, and criminal vio- modest fines and employ- exists between the ease of lations and double the upper and lower er-friendly rules of evi- employer compliance and values defining this range (Cornyn-Kyl dence. Conversely, to the the appropriate penalty and McCain-Kennedy proposals) extent that compliance for non-compliance. • Eliminate the range of possible fines, estab- can be simplified, includ- lishing flat rates of roughly double the ing perhaps through a centralized screening existing maximum fines (Tancredo proposal) system, any failure by an employer to comply fully could be treated as a substantive viola- Any increase in the penalty structure—in tion. Thus, a direct relationship exists combination with other changes discussed between the ease of employer compliance and above—would be a low-cost way to increase the appropriate penalty for non-compliance. the deterrent effect of employer sanctions. By eliminating the option to apply small fines to 6. Increase Worksite Access first-time offenders, Tancredo’s proposal in Finally, a number of steps could be taken to particular could eliminate the view that fines streamline and strengthen the enforcement are an acceptable cost of doing business. process. Regulatory hurdles identified above Legislators could also consider steps to which prohibit or delay worksite inspections increase the threat of jail time for non- could be eliminated (by statute or regulatory compliant employers. (The Cornyn-Kyl and change). A more aggressive step would be to Tancredo proposals would both penalize work- permit no-warrant worksite searches for immi- ers for lying on I-9 forms, including potential- gration violations. Access could be limited to ly with jail time.) In addition, employers’ use a particular class of worksites (as was consid- of labor contractors to avoid sanctions could ered during IRCA negotiations) or be based

11 on a probable cause finding following an audit therefore be considered: the creation of a of verification records. centralized screening system and of a job holder database. The promise of more worksite access could significantly strengthen the cases investigators Centralized Screening The current system make against non-compliant employers, but a makes employers responsible for confirming trade-off exists between this increased the work eligibility of job applicants, a system enforcement capacity and employers’ civil lib- that is analogous to requiring police to admin- erties. No-warrant searches would conflict ister driving tests on traffic stops rather than with employers’ Fourth Amendment protection relying on information contained in the dri- from undue search and seizure, and steps in ver’s license, or requiring hospitals to admin- this direction would be politically difficult and ister an anatomy quiz to prospective surgeons subject to legal challenges. Still, there may be rather than trusting their board certification. room for increasing worksite access without Under a centralized system, the responsibility overstepping these bounds, for example, for verifying work eligibility would rest with through eliminating the requirement that professional screeners at the point of docu- agents obtain written supervisor approval prior ment issuance, and proof of eligibility would to worksite inspections. be embodied in a worker’s identity card itself. Employers could V. Centralized Screening and thus assume card- Under a centralized system a Job Holder Database holders are work- the responsibility for verify- authorized, and ing work eligibility would Each of the reforms discussed above has the employer responsi- rest with professional potential to strengthen the existing employer bility would be screeners at the point of sanctions system, and taken together these reduced to keeping changes could substantially increase compli- a record of new document issuance, and ance. Yet the viability of the strategy outlined hires. Enforcement proof of eligibility would be above depends on the success of technological agents would be embodied in a worker’s responses to document fraud and on the suc- responsible for identity card itself. cessful implementation of a fast and reliable insuring the employer verification system. integrity of work authorization documents, and for analyzing employer records to search for Document reforms face an uncertain future. evidence that employees are using borrowed Data entry errors, alternative spellings, and or stolen documents. Employers would be update delays guarantee that even a reformed sanctioned for failure to maintain accurate verification database will remain imperfect— personnel records, but not for having unautho- especially in combination with a universal rized workers on their payroll per se. employer interface. Thus, eligibility verifica- tion is likely to remain deeply flawed, under- “Front-loading” verification in a centralized mining the ability of employers with good screening system would benefit workers, intentions to comply effectively and undermin- including, in particular, victims of Basic Pilot ing the ability of enforcement agents to hold false negative screening errors. On one hand, remaining employers accountable. conducting eligibility screening at the point of document issuance rather than the work- Two more fundamental reforms to the site would reduce the error rate by eliminat- employer sanctions system should also ing a weak link in the chain (employer data

12 Policy Brief

entry) and by maximizing the use of screening system. In this case, rather than machine-readable documents. On the other query an eligibility database, employers would hand, a centralized system would also allow be required to submit a worker’s ID number to the inevitable false negatives remaining to be a job holder database. This registration resolved at this early stage and in a durable process would be a minimal burden since fashion. Thus, workers would establish eligi- employers would not await a response. REAL bility on a regular schedule at the point of ID or other machine-readable cards will even- card-issuance or renewal, rather than with tually allow registration to consist of a card every change of employment; and eligible swipe similar to that now employed to convey applicants would not be denied work as a credit card information. result of database errors. The creation of a job holder database would Employers would also benefit from the elimi- improve sanctions enforcement in two impor- nation of delays in the hiring process. More tant ways, and also strengthen the overall generally, employer compliance would be immigration system: streamlined to the point of entering an • Employers’ registration of new hires employee’s unique work authorization num- would replace completed I-9 forms as ber—eventually via low-cost magnetic scan- proof of employer compliance. Electronic ning using credit card technology—and the registry would greatly simplify the process problem of unintentional non-compliance of auditing employer records and thus would be virtually eliminated. As noted above, raise conviction rates among non-compli- by radically reducing the burden of compli- ant employers. ance, a centralized screening system could be • A job holder database would outperform associated with stricter enforcement of these an eligibility verification database as a reduced responsibilities. tool for detecting identity fraud because it would more easily detect cases in which To the extent that the REAL ID Act and the same individual appears to hold mul- Cornyn-Kyl’s proposed Social Security tiple jobs simultaneously. A job holder reforms are fully implemented (or that a database would be a particularly valuable secure one-card system is established), the tool in this regard if employers were also document-issuance process would become a required to notify the database when a de facto centralized screen. Thus, establish- worker’s employment ends. ing a formal centralized screening system • A job holder database would also be an would mainly require changes in the liability asset in the management of non-immi- structure of employer sanctions, and the elim- grant workers such as current H-2A or ination of a redundant eligibility check by H-1B workers or future participants in a employers. In addition, with employers larger temporary worker program. Indeed, absolved of the responsibility for checking if temporary worker rights and responsi- applicant eligibility, the I-9 record-keeping bilities were tied to their employment, as process would be replaced by the establish- would likely be the case, a job holder ment of a job holder database. database would be almost indispensable to ensure compliance. Creation of a Job Holder Database Mandatory electronic registration by employ- These benefits would have to be balanced by ers should still play a role in a centralized economic and privacy costs. The economic

13 costs of establishing and maintaining the data- lines seem optimistic, however, and the base would be modest compared to those asso- three-year timeline for implementation of ciated with building a universal employer eli- REAL ID driver’s license provisions may gibility verification database (i.e., because a be a more realistic assumption about job holder database would not respond to user when secure documents could be made queries). However, it would still need to universally available. accommodate hiring (and perhaps termina- tion) data for over 50 million annual employ- The timeline for a transition to new iden- ment changes in the United States and to tity and verification documents also search for and report on duplicate ID informa- depends on how to handle out-of-date tion. Developing and maintaining such a data- cards. If existing cards are “grandfa- base would cost about $20 million per year thered out,” then the system of eligibility (based on a comparison with the FTC’s verification would remain insecure National Do Not Call Registry). In addition, throughout the period in which old cards while a database of job holders would be a may still be used. In this regard, a one- valuable tool for the enforcement of employer card system may allow a quicker and sanctions, it would create an additional point smoother transition than one which seeks of vulnerability which might be exploited by to update existing forms of identification cyber criminals, exacerbating the privacy and verification. issues raised above. • Create a database to handle manda- VI.Timeline for Potential Reforms tory screening. The most difficult obsta- cle to effective employer sanctions (beyond, perhaps, lack of political will) is Many of the potential reforms discussed here the problem of eligibility verification. could be implemented readily, including While verification infrastructure exists in reforms to the penalty structure and steps to the form of the Basic Pilot, whether the increase agents’ access to worksites. Other program can be enhanced enough to man- changes, like hiring additional ICE agents or age a 2,500-fold increase in its query rate requiring additional DOL collaboration, and whether it can be restructured to could be phased in over time without sub- detect identity fraud are highly uncertain. stantially disrupting the enforcement process. Even if these issues are addressed by the Others would require additional time to McCain-Kennedy proposal for construct- implement and could be disruptive during ing a new SSA/DHS database, it would the transition period. likely take several years before the new • Issue secure identification and verifi- system could be readied for universal use. cation documents. The Cornyn-Kyl bill Updating the verification database for use would mandate secure immigration and in a centralized screening system would visa documents by late 2007, while speed this process significantly by allow- McCain-Kennedy calls for quicker imple- ing for a greater level of operator expert- mentation of secure immigration and visa ise and tolerance for a slightly higher documents over a six-month period. error rate (since errors would be resolved Cornyn-Kyl would also mandate that SSA during document issuance, and a single exclusively issue secure cards within a correction process should produce a year of its passage. These one-year time- durable fix).

14 Policy Brief

• Create a job holder database. By def- This brief has discussed six sets of changes inition, a job holder database would be which could strengthen the existing sanctions created gradually as a function of employ- system: improved document integrity, docu- er participation in a mandatory registra- ment consolidation, mandatory electronic eli- tion program. Nonetheless, a database gibility verification, increased enforcement infrastructure would need to be estab- staffing and institutional commitment, a lished and equipped to handle phone-in stronger penalty structure, and increased and internet-based registrations. One worksite access by enforcement agents. In model is the National Do Not Call addition, this brief recommends two more fun- Registry, which was formally established damental changes: in December 2002 and launched six • Centralized screening. A centralized months later. As of April 2005, the reg- screening system would redefine the divi- istry had accepted 92 million phone num- sion of responsibility between employers bers, accommodating an input rate similar and enforcement agents. Professional to that required of a job holder database. screeners would establish work eligibility Magnetic scanners linking employers to at the point of document issuance, and the database could be phased in over time employers would be responsible simply as machine-readable identification cards become widespread. for keeping a record of employees hired, ideally by swiping a machine-readable • Create a new worksite enforcement document. Such a system would be more division or task force. One model for consistent with other regulatory areas and the creation of a new entity devoted would limit the impact of inevitable verifi- exclusively to worksite enforcement is the cation errors. Minimizing the burden of establishment of the Occupational Safety compliance means employer responsibili- and Health Administration (OSHA) by the ties could be more strictly enforced. December 1970 Occupational Safety and Heath Act. Within two years of the Act’s • Job holder database. A centralized signing, OSHA had assembled a staff de screening system would eliminate the novo, developed a set of safety standards, need for an employer query of an eligibil- and established its own training institute ity database, but registration of new hires for OSHA investigators. should be recorded in a job holder data- base. Such a database would provide a VII. Recommendations and targeting tool to identify both non-compli- Conclusions ant employers and migrants submitting borrowed or stolen documents. A job holder database would also strengthen In sum, the current system based on employer the auditing process during investigations screening of diverse and non-secure I-9 docu- and could play an important role in the ments, the voluntary Basic Pilot program, lim- governance of a new temporary worker ited institutional commitment to worksite program if it is enacted. enforcement, and employer-friendly enforce- ment and penalty rules utterly fails to deter Finally, this discussion raises a pair of addi- employers who prefer to hire unauthorized tional issues that deserve careful attention migrants, and provides limited tools to those prior to considering sanctions reforms: employers attempting substantive compliance with IRCA rules.

15 • The problem of successful employer • Three fundamental trade-offs. The sanctions. Roughly 11 million unautho- American political system has identified rized immigrants now reside in the preventing unauthorized migration as a United States, and the majority of them priority. Yet accomplishing this goal, are regularly employed in a wide range including through employer sanctions, of economic sectors and geographic conflicts with three other social goods: regions. Any effort to eliminate this citizens’ and legal residents’ freedoms workforce without creating a legal sub- and their right to privacy, a pro-growth stitute would wreak havoc on the US regulatory environment, and relatively economy, at least in the short run. In strict limits on legal immigration. In addition, eliminating employment oppor- short, the only way to achieve greater con- tunities would cause hardship for mil- trol over unauthorized immigration is to lions of individuals and their families give ground in one or more of these other living within the United States. Thus, areas. While the post-9/11 period has any meaningful reforms to the employer been characterized by unusual tolerance sanctions system should be preceded by for limits on civil rights, in the long run, reforms to address this population and resolving legal migration supply and ongoing employer needs. Care should demand tensions would facilitate effective also be taken to ensure that employers control of unauthorized flows at a far do not “spread the risk” associated with lower cost. a more effective sanctions regime by cut- ting workers’ wages, a tendency observed following IRCA’s passage.

Author’s Note

Several additional bills were introduced in the 109th Congress after the September 21 Task Force meeting for which this brief was prepared; they are not discussed here. Additional relevant legislation includes: the Enforcement First Immigration Reform Act of 2005 (HR. 3938), intro- duced by Representative J.D. Hayworth (R-AZ); and the Employment Verification Act of 2005 (S. 1917), introduced by Senator Chuck Hagel (R-NE).

At the time this brief went to press, the Senate Judiciary Committee was preparing to mark up a comprehensive reform bill which includes elements drawn from the Cornyn-Kyl, McCain- Kennedy, and Hagel proposals.

Acknowledgements

The author would like to thank David A. Martin and Doris Meissner for valuable advice and guid- ance on this project.

16 Policy Brief

Additional Readings

Brownell, Peter, “The Declining Enforcement of Government Accountability Office, Employer Sanctions,” Migration Information “Telemarketing: Implementation of the Source. September 1, 2005. See www.migrationin- National Do Not Call Registry” (GAO-05-113). formation.org/Feature/display.cfm?ID=332. Washington, DC: US Government Accountability Office. 2005. Commission on Immigration Reform, U.S. Immigration Policy: Restoring Credibility. Government Accountability Office, Washington, DC: US Government Printing “Immigration Enforcement: Weaknesses Office. 1994. Hinder Employment Verification and Worksite Enforcement Efforts” (GAO-05-813). Government Accountability Office, “Social Washington, DC: US Government Security: Mass Issuance of Counterfeit-Resistant Accountability Office. 2005. Cards Expensive, but Alternatives Exist” (GAO/HEHS-98-170). Washington, DC: US Jernegan, Kevin, “Verifying Employment Government Accountability Office. 1998. Eligibility.” Background paper prepared for the Independent Task Force on Immigration and Government Accountability Office, “Illegal America’s Future, Migration Policy Institute. Aliens: Significant Obstacles to Reducing September 2005. Unauthorized Alien Employment Exist” (GAO/GGD-99-33). Washington, DC: US Martin, Philip and Mark Miller, “Employer Government Accountability Office. 1999. Sanctions: French, German, and US Experiences,” International Migration Papers #36. Geneva: International Labour Office. 2000. List of Task Force Members Co-Chairs: Members:

Spencer Abraham, Chairman and CEO, The T. Alexander Aleinikoff, Dean of the Law Abraham Group, LLC; Distinguished Visiting Center and Executive Vice President for Law Fellow, Hoover Institution, former Secretary of Center Affairs, Georgetown University; former Energy; and Senator (R) from Michigan General Counsel, United States Immigration and Naturalization Service (INS) Lee Hamilton, President and Director, Woodrow Wilson International Center for Howard Berman, (D) Member of Congress, Scholars; former Vice Chair, 9/11 Commission California and Member of Congress (D) from Indiana Jeanne Butterfield, Executive Director, Ameri- Director: can Immigration Lawyers Association (AILA)

Doris Meissner, Senior Fellow, Migration Policy Oscar A. Chacón, Director, Enlaces América, Institute; former Commissioner, United States Heartland Alliance for Human Needs and Immigration and Naturalization Service (INS) Human Rights

17 Thomas J. Donohue, President and CEO, Frank Sharry, Executive Director, National United States Chamber of Commerce Immigration Forum

Jeff Flake, (R) Member of Congress, Arizona Debra W. Stewart, President, Council of Graduate Schools; former Vice Chancellor and Fernando Garcia, Executive Director, Border Dean of the Graduate School, North Carolina Network for Human Rights State University.

Bill Ong Hing, Professor of Law and Asian C. Stewart Verdery, Jr., Principal at American Studies, University of California, Davis Mehlman Vogel Castagnetti, Inc., and Adjunct Fellow, Center for Strategic and International Tamar Jacoby, Senior Fellow, Manhattan Studies (CSIS); former Assistant Secretary, Institute Department of Homeland Security (DHS)

Juliette Kayyem, Lecturer in Public Policy at John Wilhelm, President, Hospitality Industry the John F. Kennedy School of Government at of UNITE HERE Harvard University; former member of the National Commission on Terrorism James W. Ziglar, former Commissioner, United States Immigration and Naturalization Edward Kennedy, (D) Senator, Massachusetts Service (INS)

John McCain, (R) Senator, Arizona Ex Officio

Janet Murguia, President and CEO, National Thor Arne Aass, Director General, Council of La Raza Department of Migration, Ministry of Local Government and Regional Development, Norway Leon Panetta, Director, Leon and Sylvia Panetta Institute for Public Policy, California Malcolm Brown, Assistant Deputy Minister, State University at Monterey Bay; former Chief Strategic Policy and Communications, of Staff to the President; former Director, Office Citizenship and Immigration Canada of Management and Budget (OMB) Jean Louis De Brower, Director, Immigration Steven J. Rauschenberger, Senator, State of and Asylum, European Commission Directorate Illinois; President-elect, National Conference of General for Justice, Freedom and Security State Legislatures; Deputy Republican Leader and former Chairman, Illinois Senate Geronimo Gutierrez, Undersecretary for Appropriations Committee North America, Ministry of Foreign Affairs, Mexico Robert Reischauer, President, Urban Institute; former Director, Congressional Budget Representatives of the Department of Homeland Office (CBO) Security, Department of Labor, and the Domestic Policy Council of the White House Kurt L. Schmoke, Dean, Howard University School of Law; former Mayor, Baltimore, MD

18 Policy Brief

About the Author Marc R. Rosenblum Marc R. Rosenblum is an International Affairs Fellow at the Council on Foreign Relations and a Visiting Scholar at the Migration Policy Institute. Dr. Rosenblum is an Assistant Professor of Political Science at the University of New Orleans, where he teaches courses on US-Latin American relations, Latin American and Comparative Politics,The Politics of Economic Development, and Political Methodology. Dr. Rosenblum’s monograph, The Transnational Politics of U.S. Immigration Policy, was published in 2004 by the University of California – San Diego Center for Comparative Immigration Studies. He has also published over a dozen scholarly articles and book chapters, including in the journals Annual Review of Political Science, Comparative Political Studies, Human Rights Review, Journal of Peace Research, Latin American Politics and Society, Migraciones Internacionales, and Political Power and Social Theory. Dr. Rosenblum earned his PhD in political science from the University of California – San Diego and his BA from Columbia University. Additional Publications Prepared for the Task Force Independent Task Force on Immigration Documentation Provisions of the Real and America’s Future: The Roadmap ID Act Policy Brief by Michael Fix, Doris Meissner, and Backgrounder by Kevin Jernegan Demetrios G. Papademetriou Reflections on Restoring Integrity to the An Idea Whose Time Has Finally Come? United States Immigration System: A The Case for Employment Verification Personal Vision Policy Brief by Tamar Jacoby, Manhattan Institute Insight by Demetrios G. Papademetriou

US Border Enforcement: From Unauthorized Migrants: Numbers and Horseback to High-Tech Characteristics Insight by Deborah W. Meyers Report by Jeffrey S. Passel, Pew Hispanic Center

Eligible to Work? Experiments in Twilight Statuses: A Closer Examination of Verifying Work Authorization the Unauthorized Population Insight by Kevin Jernegan Policy Brief by David A. Martin, MPI and the University of Virginia Immigration Enforcement Spending Since IRCA Lessons from the Immigration Reform Fact Sheet by David Dixon and Julia Gelatt and Control Act of 1986 Policy Brief by Kevin O’Neil and Betsy Cooper IRCA’s Employer Sanctions: Legal Status as a Labor Standard The “Regularization” Option in Managing Backgrounder by Julia Gelatt Illegal Migration More Effectively: A Comparative Perspective Policy Brief by Demetrios G. Papademetriou

19 1400 16th Street NW Suite 300 Washington, DC 20036

202 266 1940 202 266 1900 (fax) www.migrationpolicy.org

www.migrationinformation.org

www.migrationpolicy.org

Future, please visit: please Future,

www.migrationinformation.org. For more information on the Independent Task Force on Immigration and America’s and Immigration on Force Task Independent the on information more For

nomto orewbst,at site, web Source Information this project. this

Manhattan Institute and the Woodrow Wilson International Center for Scholars for Scholars for Center International Wilson Woodrow the and Institute Manhattan

produces the Migration the produces

icated to the analysis of the movement of people worldwide, is partnering with partnering is worldwide, people of movement the of analysis the to icated

vrmr nertdwrd MPI world. integrated more ever groups; and public policy and immigration experts. MPI, a nonpartisan think tank ded- tank think nonpartisan a MPI, experts. immigration and policy public and groups;

who are involved in shaping legislation; leaders from key business, labor and immigrant and labor business, key from leaders legislation; shaping in involved are who

ties that migration presents in an in presents migration that ties

The approximately 25 task force members include high-ranking members of Congress of members high-ranking include members force task 25 approximately The

to the challenges and opportuni- and challenges the to

Service.

demand for pragmatic responses pragmatic for demand Doris Meissner, the former Commissioner of the Immigration and Naturalization and Immigration the of Commissioner former the Meissner, Doris

(D-IN) serve as co-chairs, and the task force’s work is directed by MPI Senior Fellow Senior MPI by directed is work force’s task the and co-chairs, as serve (D-IN)

levels. It aims to meet the rising the meet to aims It levels.

Former Senator Spencer Abraham (R-MI) and former Congressman Lee Hamilton Lee Congressman former and (R-MI) Abraham Spencer Senator Former

the local, national, and international and national, local, the

ommendations in 2006. in ommendations

migration and refugee policies at policies refugee and migration The panel’s series of reports and policy briefs will lead to a comprehensive set of rec- of set comprehensive a to lead will briefs policy and reports of series panel’s The

development, and evaluation of evaluation and development,

Integrating immigrants into American society American into immigrants Integrating ■

Labor markets and the legal immigration system immigration legal the and markets Labor The institute provides analysis, provides institute The ■

Immigration enforcement and security requirements security and enforcement Immigration

■ movement of people worldwide. people of movement

The growing unauthorized immigrant population immigrant unauthorized growing The

■

dedicated to the study of the of study the to dedicated

The task force’s work focuses on four major policy challenges: policy major four on focuses work force’s task The

partisan, non-profit think tank think non-profit partisan,

information and workable policy ideas. policy workable and information

leaders from key sectors concerned with immigration, which aims to generate sound generate to aims which immigration, with concerned sectors key from leaders MI sa needn,non- independent, an is (MPI)

Immigration and America’s Future. The task force is a bipartisan panel of prominent of panel bipartisan a is force task The Future. America’s and Immigration

The Migration Policy Institute Policy Migration The

This report was commissioned as part of MPI’s Independent Task Force on Force Task Independent MPI’s of part as commissioned was report This MORE FROM MPI: FROM MORE