The right to freedom from forced labor is firmly established in international human rights law. Core conventions of the International Labor Organization dating back to 1930 require the abolition of forced or compulsory labor, which the conventions define as work "exacted.under the menace of any penalty and for which the [worker] has not offered himself voluntarily." ^[1] The 1948 Universal Declaration of Human Rights provides that "everyone has the right . . . to free choice of employment . . . ." ^[2] Codes of conduct often specifically identify prison labor, indentured labor, and bonded labor as among the forms of forced labor that are prohibited. ^[3]

Debt-bonded labor in various forms is a particularly pervasive problem, especially for migrant workers who bond themselves - often to middlemen - as their only means of reaching distant jobs.As with overtime work hours, the question of voluntariness looms large in trying to measure whether involuntary labor is taking place.Recorded cases of slave-like working conditions under debt bondage are common and include practices such as beatings and food deprivation, confiscating passports and return plane tickets, and imposing exorbitant additional debts on arrival that were not part of the original agreement, in addition to forcing work with no pay. ^[4] Such cases occasionally reach law courts that can force corrective action. ^[5]

Although voluntariness is difficult to pin down directly withunits of measurement, current measurement practice in this area is capable of producing some unusually reliable indicators for whether involuntary labor, at least in its starker forms, is occurring in a particular workplace.  

Measurability issues 

The term "involuntary labor" covers a wide spectrum of situations, from outright slavery to debt bondage to a fear that overtime work cannot be safely declined. Units of measurement need to cover the full spectrum. Some forms of involuntary labor can be established primarily with objective evidence, while others require the measurement of subjective states of mind. 

The simplest defining question for involuntary labor is: can the worker quit working when he or she wants to? A constraint like physical confinement of workers, or holding a worker's legal documents hostage, is relatively straightforward to measure. But economic constraints, though just as powerful, are harder to gauge for involuntariness. Where does the normal economic pressure any worker feels to get and keep a job, shade into abusive economic pressure to deprive the worker of his or her freedom?The more economically desperate the worker, and the fewer employment alternatives he or she has, the harder these factors are to separate. 

One subset of involuntary labor is coerced overtime, which is also treated in Working Hours.The simple defining question in that case is not whether the worker can quit her job, but whether she can limit her job to its proper hourly boundaries. 

In terms of current measurement practice, this area of workers' rights (along with Abuse and Harassment) shows the smallest number of measurements, as well as the fewest that go beyond the yes/no form.But unlike the area of Abuse and Harassment, existing measurement units in the area of involuntary labor provide the basis for some very useful indicators of compliance with or violation of code standards. 

Current strengths

Debt bondage and physical restrictions on freedom of movement are the best-measured indicators of involuntary labor in current measurement practice. 

Debt bondage. Because debt to the employer usually leaves a paper trail, debts of the size that in practice amount to debt bondage are relatively easy to measure, as are the bookkeeping practices and employer policies that encourage or discourage it. Yes/no questions about such practices and policies are more useful than in other contexts, because the answers can be checked in an audit of financial records when workers identify debt bondage as a problem. The incentive to create false records is also less in this context than in the context of wages (see Wages), since to hide debt bondage through falsified records, the false record would have to be in the worker's favor (i.e., a record created by the employer showing that no debt from the worker to the employer was owed).Debt bondage to middlemen is harder to trace, but even middlemen tend to rely on repayment directly from the employer, which the employer's books can be audited for. 

Step one in auditing pay records for debt bondage is to determine if deductions are being made from wages in order to repay recruitment fees. ^[6] A strong anti-bondage indicator exists if the employer itself pays recruitment fees ^[7] or reimburses the employee for them, ^[8] either way keeping a potential debt burden for recruitment fees from falling on the employee. 

Short of an employer-pays rule, honest fiduciary bookkeeping on behalf of the employee is a next-best alternative.Useful units of measurement for honest bookkeeping in this area include keeping itemized statements from employment brokers, ^[9] disclosing the identity of all brokers used by the employer, ^[10] and barring any employment contracts with third parties. ^[11] Making sure that brokers are informed of code provisions, and of the employer's obligation to comply with them, is also a useful (if only partial) deterrent for both brokers and their employer clients. ^[12]  

Not specific to recruitment fees, but a practice that strongly indicates a risk of debt bondage, is the requiring of a salary deposit ^[13] (e.g., of a month's wages ^[14] ) that the employer holds during the period of employment. Although the deposit is supposed to be returned when employment is completed, the employer's control of the deposited funds can make it difficult for employees to leave when there is temporarily no work. ^[15] An employer using control of the deposit to inhibit departure or resignation ^[16] is a clear indicator of an involuntary-labor practice. More generally, any program of company loans to employees can be probed for fairness and proportion. ^[17]  

Physical constraints.Whether workers are free to leave the workplace ^[18] or are locked in, for example until production quotas are met, ^[19] is a basic indicator of involuntariness that is relatively easy to observe or learn about through questions to workers themselves ^[20] or local observers. ^[21] More complex questions arise in the context of security guards, who may be just as powerful a constraint on workers' freedom of movement as a locked gate but whose intimidation effect may be harder to gauge. ^[22] Checking for the presence of security guards ^[23] and the existence of clear contractual restrictions on them ^[24] are preliminary if not definitive units of measurement. 

Constraints on workers' freedom in non-work hours are as important as constraints in the workplace itself. The common practice of the employer holding the employee's identity papers, or other documents needed for off-site movement, is an important indicator of involuntariness both in the workplace and during off hours. ^[25] Employers that claim to need the documents can keep copies rather than originals, ^[26] and employers claiming to be safekeeping the documents on workers' behalf can be checked for how easily workers can get their documents on demand. ^[27]  

The other focus of attention for off-hours constraints is dormitory housing. Security personnel can have the same inhibiting effect in dormitories as in the workplace, ^[28] and visitor policy ^[29] and practice ^[30] can be also be checked. So can residents' access to telephones in dormitory housing. ^[31]  

Migrants often face constraints in the workplace that non-migrant workers do not. Not only are they more likely to have difficulty with travel documents and dormitory restrictions, as well as language and cultural obstacles with co-workers and managers, but they face an ever-present threat, implicit and often explicit, that they will lose their status as legal migrants and be deported if they complain about any mistreatment (in the process, probably forfeiting any outstanding pay that is due to them). Comparing migrant and non-migrant worker conditions ^[32] in the workplace is a good preliminary indicator for migrant-specific problems of involuntariness, although not enough by itself to establish violations. 

Current weaknesses 

Prison Labor.Although the use of prison labor in factories is perhaps the most blatant form of involuntary labor practice, it is very poorly measured in current practice, with only the most general of queries to monitors ^[33] or local NGOs ^[34] and local unions, ^[35] usually in a form that requires the monitor to decide what "prison labor" and "forced labor" mean without any meaningful guidance. ^[36] Workers themselves are only vaguely consulted with no specific focus on prisoners, ^[37] even though non-prison workers would almost certainly know if there were prisoners working in the same factory and would have an incentive to reveal them.  

The lack of focus on prison labor is illustrated not only by the lack of careful units of measurement, but also by an attempt to displace responsibility by relying on whether the buyer company has already checked for it ^[38] - a far cry from the prevailing approach to other forms of involuntary labor (see Strengths above) or to other important workers' rights issues.Prison labor may not be as prevalent as debt bonded labor in the factories in question, but it is a serious enough abuse to warrant equally reliable measurement. Measuring whether a factory uses unpaid workers - or workers paid only through a third party - is not significantly harder than measuring other forms of involuntary labor (compare Strengths above) or other forms of wage cheating (see Wages), and would be likely to reveal any prison labor or slave labor that exists. 

There is also little focus in current measurement practice on migrant workers, even though problems of involuntariness are likely to show up most severely and blatantly among migrants. Measurement units that address confinement to dormitories, withholding of travel documents, and bonded debt do not distinguish migrants from other workers, or try to identify differentials in treatment for migrants. There are no measurement units for the potential threat of deportation that overhangs migrant workers, even if explicitly voiced by managers, although that threat can be highly coercive and can mask migrant-specific abuses. Migrant laborers should be a key target for separate units of measurement, and their treatment could be used as a strong indicator for involuntariness problems generally. 

Finally, the subjective dimension of involuntariness gets relatively little attention, even though the worker's perception about the voluntariness of her labor is arguably as important as the presence or absence of objective constraints. (See discussion of voluntary overtime in Working Hours, and discussion of worker perception in Abuse and Harassment.)Even debt bondage includes an important subjective element; i.e., whether the size of the debt is large enough to make a worker feel involuntarily bound. This element could be measured, for example by polling workers about what level of debt would make them feel that way. 

Best current practices (selected) 

·        Measuring factory policies and bookkeeping practices that can minimize debt bondage risks (see Strengths above)

·        Checking whether resignations are accepted with unquestioned return of salary deposits and workers' legal papers ^[39]

·        Emphasizing transparency in any transaction with an employment broker ^[40] and checking avoidance of any employment obligations involving third parties ^[41]

·        Establishing baseline statistics on the number of migrant workers and where they come from, so that significant changes in the proportion of migrant workers will be detected and investigated further ^[42]

·        Discovering whether migrant employees can count on the employer to return them to their home region or country ^[43] - a key indicator of involuntariness if the answer is "no" and they would be stranded if they quit

·        Scrutinizing the contractual obligations and boundaries on security guards ^[44] and making sure those contractual provisions are transparent ^[45] - not because contractual provisions accurately describe reality, but because they can have a partial deterrent effect, particularly if they are known to workers and monitors. 

Potential improvements suggested by analysis 

1.      Focus more on the presence or absence of deterrents to involuntary labor: what are the penalties for the employer if prison labor is discovered? For personnel managers? For security guards if they block egress, bully workers into forced overtime, or restrict access to dormitories by workers or their guests more stringently than provided in written rules? For paymasters if deposits or debt repayments are improperly withheld from wages?How clearly are the boundaries of acceptable behavior spelled out, so that abuses are easy to identify?

2.      Build a foundation of data on migrant workers (their relative percentage of the workforce over time, as the workforce expands and contracts; their relative pay; their relative presence in employer-sponsored housing; their relative levels of debt to the employer), separately for international migrants and domestic migrants, and use it to spot potential problems. Look for differential treatment of migrant workers in other areas of workers' rights, as an indicator of the potential for conditions of involuntary labor.

3.      Survey workers more carefully for their views on what would constitute an economic barrier to voluntary resignation (e.g., what level of debt to employer), and on whether such barriers currently exist or have existed in the factory.

4.      Cross-check more systematically with local outsiders for hints of involuntary labor practices (likely to be more successful in plants with local workers rather than plants with all or mostly migrant workers).

5.      Assemble data from multiple factories on what constitutes debt levels and debt repayment terms that amount to debt bondage, and use them to design generic measurement units (e.g., "does any worker owe the employer more than the equivalent of XX weeks of that worker's wages?").

6.      Use indirect indicators of the potential for prison labor, such as the distance of the factory from the nearest prison; whether security personnel are assigned to specific groups of workers; separate meals/isolated areas of the factory for specific groups of workers; special procedures for specific groups of workers on use of time clock, transportation, etc.