Introduction and Overview
Concerns about conditions in the workplace emerged at the very beginning of the industrial revolution. With increased international economic interdependence over the past half century, these concerns became more global, and in the late 1990s a consensus developed, as reflected in declarations by the United Nations in 1995, the World Trade Organization in 1996, and the International Labour Organization (ILO) in 1998, that certain standards—freedom of association, the abolition of forced labor and child labor, and nondiscrimination—are fundamental rights to which workers everywhere are entitled. We refer to these as the core labor standards.
The treatment of workers is embedded in broad social, political, and economic circumstances. In the absence of political will and effective governance institutions that are based on the rule of law, there is little chance that the core labor rights will be respected. In the absence of strong economic growth, there is little chance of achieving broad-based improvements in workers’ standards of living. Thus, the economic and political context is crucial when assessing compliance with labor standards, particularly for identifying the sources of noncompliance and, therefore, the appropriate policies, including technical and financial assistance, to address noncompliance. At the same time, monitoring compliance with labor standards poses challenges. The information required is often unavailable or of poor quality, and the interpretation of what the information means in terms of compliance is not always obvious.
COMMITTEE CHARGE AND FOCUS
Recognizing these difficulties, the U.S. Department of Labor (DOL) requested guidance from the National Research Council (NRC) of The National Academies. The request calls on the NRC to “…conduct a set of connected tasks designed to provide the federal government with expert science-based information on issues pertaining to international labor standards, monitoring, and enforcement.” Specifically, DOL charged the NRC to:
Identify relevant, valid, reliable, and useful sources of country-level data on labor standards and incorporate them into a database tailored to the current and anticipated needs of the Department of Labor’s Bureau of International Labor Affairs (ILAB).1
Assess the quality of existing and potential data and indicators that can be used to systematically monitor labor practices and effectiveness of enforcement, in order to determine compliance with national labor legislation and international standards.
Identify innovative measures to determine compliance with international labor standards on a country-by-country basis and measure progress on improved labor legislation and enforcement.
Explore the relationship between labor standards compliance and national policies relating to human capital issues.
Recommend sustainable reporting procedures to monitor countries’ progress toward implementation of international labor standards.
The DOL specified that the committee’s work be based on the consensus core standards identified in the 1998 Declaration of Fundamental Principles and Rights at Work of the ILO:
freedom of association and the effective recognition of the right to collective bargaining;
the elimination of all forms of forced or compulsory labor;
the effective abolition of child labor; and
the elimination of discrimination in respect of employment and occupation.
In addition, the committee was asked to consider how to assess compliance with standards related to acceptable conditions of work, including hours of work, wages, and occupational safety and health.
The first two points of the committee’s charge require comment. The report and database structure constitute the committee’s response to the first point—the identification of relevant and useful data for assessing and monitoring international labor standards. However, the sheer volume of data we so identified precluded an assessment of the quality (including reliability and validity) of the hundreds of thousands of pieces of data that we recommend be included in the database. With regard to the second point, the committee interpreted its charge with regard to quality in terms of the kinds of data that would be necessary to assess and monitor compliance with labor standards. This interpretation followed tense committee deliberations and our findings from a workshop held on July 10, 2002.2 The workshop enabled the committee to explore with a wide range of experts, the complexities involved in defining and measuring quality. We concluded that the only way to interpret the charge with respect to validity and reliability was to include all available and relevant data, along with appropriate caveats concerning inferences, interpretations, and uses of the information.
Furthermore, the committee did not have the resources to evaluate the huge quantity of data that now exist and might be considered relevant to the committee’s task, nor to devise a system to evaluate data that will become available in the future. For example, the ILO’s database of legal documents (NATLEX) alone has 55,000 pieces of information, some of which are in languages other than English. Moreover, the committee believes that assessment of the quality of any given data is an essential part of the overall process of assessing compliance.
The committee’s work is only the first step in what must be on ongoing process of identifying appropriate data, assessing the quality of those data, and then analyzing their relevant within the overall task of monitor-
ing compliance with international labor standards. This report and its accompanying database are meant to provide the U.S. DOL with a structure and procedure for carrying out that task over many years. Our recommendations point to specific ways that the task can be informed with better data, more systematic analysis, and closer working relationships among the many organizations, agencies, and individuals—in both the United States and other countries—who are involved in labor standards.
In approaching its charge, the NRC Committee on Monitoring International Labor Standards understood that the underlying rationale for its work is to inform U.S. policy as it relates to improving the lives of workers and their families around the world. The committee undertook the task of providing science-based information on the monitoring and enforcement of international labor standards while recognizing that compliance with core labor standards is one prong of what must be a multifaceted strategy toward achieving the fundamental goal of raising living standards across the globe.
From a historical perspective, the committee is aware that it took many decades of economic, social, and political development—including enactment and enforcement of laws on education and workers’ rights and the development of trade unions—to mostly eliminate child labor and achieve the 40-hour work week and other improvements in contemporary industrialized countries (see, e.g., Engerman, 2003). Indeed, full enjoyment of the core labor standards is an ongoing struggle everywhere in the world. This is true even though the world is far different in the twenty-first century than it was in the nineteenth. Today, democracy and free markets are far more widely spread than they were 100 years ago, and institutions such as the ILO exist to assist countries in improving labor standards if they lack the capacity themselves.
The committee is also aware that there is small likelihood of achieving broad-based, sustainable improvements in workers’ standards of living without solid economic growth, expanded market access for developing country products, and reliable governing institutions backed by respect for rule of law. Yet the empirical evidence shows that the core standards can be improved even in poor countries if the political will exists to do so. For example, some poor countries have much lower rates of child labor than
other countries at roughly the same per capita income level, while some middle-income countries have far higher rates than others (Elliott, 2004).
Although cognizant of the myriad factors that affect political, economic, and social development, and the connections between those and working conditions, the committee understood its charge as more narrowly focused on the development and analysis of information to monitor and assess country-level compliance with core labor standards and acceptable conditions of work. Thus, in its work, the committee considered only a small part of the relevant academic literature. The committee also did not delve into the historical evolution of labor standards in the now developed countries, in part, because, the economic, technological, and institutional environment facing developing countries today is quite different.
The committee also does not address in detail the disparities in working conditions among sectors in the currently poor countries, though we are well aware of the fact that the available data typically reflect conditions only in parts of the formal sector. Since the majority of people in poor countries typically work in agriculture and the informal sector and since conditions in these sectors are typically worse than those in the formal sector, the available data tend to understate the gap between actual conditions and international norms. Moreover, increased efforts at compliance with labor standards will usually not reach the majority of workers because, by definition, their activities are largely unregulated. We do not analyze these issues in depth, but we do recommend that additional effort be put into data collection and research on working conditions in the informal sectors.
As policy makers and others implement the database system proposed in this report, they need to remain cognizant of the broader social, economic, and political context in which compliance with labor standards takes place. They also need to remain open to emerging research findings that could influence both the refinement of the database system and its use.
Chapters 2 and 3 describe and evaluate the sources of information available to assess the indicators of compliance with the four core labor standards and acceptable conditions of work. Chapter 2 discusses the problems of finding qualitative and quantitative evidence about compliance that is objective, representative, and comparable over time. Chapter 3 examines
the opportunities and challenges posed by “voluntary” nongovernmental monitoring and reporting.
Chapters 4-7 address the assessment of compliance with each of the four core international labor standards. Each chapter discusses the challenges of defining, implementing, and measuring compliance with a particular standard and offers a close look at relevant sources of evidence. Chapter 8 is devoted to the assessment of acceptable conditions of work, following the pattern of Chapters 4-7 on the core labor standards.
Chapter 9 considers the connection between investment in human capital and observance of labor standards.
Part of the committee’s task was to construct a database for use in monitoring international labor standards. In response to this part of its charge, the committee has constructed a database structure, which resides on a website, WebMILS. The design of the database follows the committee’s work in this volume. There is a cell for every political jurisdiction for every indicator (see Chapters 4-8) for the four core labor standards and acceptable conditions of work, divided into three areas: legal framework, government performance, and overall outcomes. The committee also included associated factors that might either provide contextual information or serve as useful signals that there may be problems with compliance.
We stress that WebMILS is a structure: it is populated with data for only one country, to provide an example. It will be the task of the U.S. DOL to maintain, add to, and improve the database over time. The database structure is designed to include information from all possible sources on the indicators that the committee has determined are relevant to monitoring international labor standards. In addition to the data and their sources, the database is designed to include comments on anything that is known about the data or their sources (such as frequency or range of coverage). Thus, the database structure provides a way for all assessors to have access to all available data and information about those data. In this way, we have tried to devise a system that encourages dialogue among all parties and that stimulates a continuous flow of ideas about how the process of promoting and ensuring compliance can be improved.
Appendix A provides a guide to the database and sample views to enable readers to have a clear view of the committee’s product.
THE ILO AND CORE LABOR STANDARDS
While the broad principles behind the four core international labor standards are fairly clear, what they mean in practice is not always obvious. In order to flesh out the meaning, the committee believes that it is necessary to begin with the work of the ILO and to review how the ILO interpretation of the concept of “core labor standards” has evolved. The ILO has 175 member nations. Its unique tripartite structure also includes representatives from national organizations of workers and employers.
Traditionally, the ILO has established international labor standards by formulating conventions, which are multilateral treaties that are binding on the countries that ratify them. Once a country ratifies a convention, it has an international obligation to implement the treaty in law and practice. To hold ratifying countries accountable for meeting their international obligations, the ILO has established a sophisticated supervisory system that includes ongoing reporting, dialogues with member nations, general complaints procedures, and a special procedure specifically for complaints related to freedom of association. This system is a key source of information on compliance with international labor standards, much of it available on the Internet (see Chapter 2).
In 1995, the ILO identified four basic rights as “fundamental to the rights of human beings at work” and launched a campaign to increase the ratification of the conventions corresponding to these rights. The basic rights are freedom of association (87) and the right to organize and bargain collectively (98), freedom from forced labor (29, 105), equal remuneration (100) and nondiscrimination in employment (111), and the abolition of child labor (138). An additional child labor convention—addressing the “worst forms” (182)—was passed in 1999. As of October 2003, 75 percent of the 175 ILO member nations had ratified seven or more of these eight conventions (International Labour Organization, 2003). These eight conventions comprise the core labor standards.
In 1998 the ILO adopted a Declaration on Fundamental Principles and Rights at Work, committing ILO member nations to realize and achieve at the policy level the four basic rights as an obligation inherent in ILO membership, regardless of whether or not they have ratified conventions
corresponding to those rights. The declaration addresses the principles and rights that are the subject of the eight “fundamental” conventions, but signing the declaration does not obligate a member government to carry out the detailed legal provisions of the conventions.
The committee’s starting point for interpreting the meaning of the four core international labor standards is the jurisprudence built up over the years by ILO experts as it relates to the eight fundamental conventions. The committee recognizes that elements of this jurisprudence remain controversial and unsettled, but there is no alternative consensus definition and, in general, it was not our role to rewrite the meaning of these standards.
As requested by the DOL, the committee also analyzes how to assess compliance with “acceptable conditions of work” (Chapter 8). This set of standards is more problematic because wages, hours, and some other conditions of work necessarily vary with average per capita income levels. Currently, there is no broad international consensus on how to apply these standards globally.
Challenges Assessing Compliance
The committee viewed its primary charge as organizing and structuring data from multiple sources into a useful tool for policy makers (and others) who wish to assess compliance. The first observation must be that problems abound. First, with ILO jurisprudence, there are difficulties with defining what the core labor standards mean in practice. Second, there are difficulties in identifying operational indicators of compliance with core labor standards so that an observer easily recognizes compliance or noncompliance. Third, there are difficulties with separating intention from capability in government performance in implementing compliance and with evaluating government effort in relation to scarce resources and competing needs for resources. Fourth, there are difficulties with finding accurate, representative, comparable sources of information about compliance or noncompliance. In particular, much of the data currently available from developing countries reflects conditions in the more modern manufacturing sector and is not representative of conditions in agriculture and the informal sector. Fifth, there are problems in drawing valid inferences from the information sources that are available. Sixth, assessments of changes in conditions over time are difficult because definitions and sampling methods
may change over time. These difficulties are detailed throughout the rest of the report.
To take one case in point, what appears to an American in 2003 as a long workweek, say, 50 hours, in large manufacturing firms in some poor countries is less than that which characterized American workers in the early twentieth century (Jones, 1963), when the United States was already much richer (in terms of per capita income) than currently poor countries. In 1890, children’s maximum hours of work in Belgium were 72 and in Denmark 63 (Huberman, 2002). There is an extensive literature on the evolution of working hours, one of the “acceptable conditions of work,” that shows its relatively close relation to per capita gross domestic product (GDP). In today’s developing countries a typical workweek for agricultural workers or those in urban workshops may be 70 hours. Recent studies in countries such as Vietnam find that the number of hours worked on farms by children decrease systematically with improvement in parental income. A 50-hour workweek in manufacturing is thus considerably less onerous than the conditions under which a majority of the population work. In turns, these long hours are one of many unfortunate effects of living in a poor country. It is an effect, not a cause, of poverty.
As yet another example, consider average earnings, a suggested measure. Assume that the available measure is earnings per year. In the United States, this is a well-defined concept, and some confidence can be placed in the reported figure. Nevertheless, even in the United States, part of the movement in earnings per year reflects movements in the business cycle and changes in hours worked—that is, changes that reflect larger macroeconomic forces that affect the labor market. In developing countries, these effects are often much stronger. As noted above, a large part of the labor force is employed in agriculture and the informal sector; earnings in these sectors are not covered adequately by most statistical offices and are subject, if available, to huge sampling errors. Movements in earnings may reflect the effects of such events as droughts and civil wars, as well as policy-induced changes, such as the unavailability of fertilizers. Thus, an improvement in annual earnings may simply reflect the termination of a civil war or improved policies that permit the import of fertilizer rather than changes in labor market policies.
In sum, the problem faced by policy makers working in this area is similar to the problem faced in other areas: integrating raw, qualitative, and quantitative data from multiple sources of differing reliability into valid and fair judgments about performance. In this case, policy makers will have
to rely relatively heavily on qualitative sources of information, including legal documents, expert (e.g., ILO) assessments, complaints, investigations of complaints, and other kinds of qualitative reports.
The committee considered, at one extreme, the possibility—often tempting—that we simply draw up an exhaustive list of the difficulties with definition, measurement, and evaluation, compounded by the troubles with data quality and inference, and turn the results over to the DOL, and other assessors, and wish them well in their endeavors. At the other extreme, many who consulted with the committee urged us to devise a detailed scoring system to rate and rank countries in terms of their compliance with each of the core labor standards.
The committee has chosen a middle road. We see our charge as trying to provide the DOL, and other assessors, with a roadmap for assessing compliance and evaluating progress or deterioration over time. The goal is to improve on the existing process by pointing the way to better information collection and dissemination and by recommending that the decision-making process itself be made more open and transparent. The list of assessors that might find our efforts useful includes government officials, elected representatives, international organizations, socially responsible corporations and investors, board members, stockholders, labor organizations, and a wide range of nongovernmental organizations.
Overall the task of assessing compliance may resemble a judge in a courtroom forming a judgment about what exactly a given labor standard requires in particular circumstances and evaluating performance based on a preponderance of the evidence (some of which is admissible, some of which is questionable, and some of which must be rejected)—more than an econometrician running regressions or a physicist conducting an experiment under controlled conditions. While the goal of the committee is to show how assessments might be made more fair and objective and balanced, in a transparent process in which alternative assessors can challenge each other’s evaluations in a clear, precise, and thorough fashion, the committee does not intend to leave the impression that there is a scientific procedure by which assessors can use science-based information to come to a scientific conclusion about compliance across all countries and all time periods.
To comply with the request from the DOL, we have developed a set of potential indicators of compliance and a detailed database structure that lists the information needed. We believe this database, when the cells are full, can provide a basis for assessing compliance. In the interim, we believe
the database structure provides information on what additional data need to be collected.
AN APPROACH TO ASSESSING COMPLIANCE
Three Sets of Indicators
WebMILS, the database system developed by the committee, contains key indicators we have identified to help in assessing compliance. Each of the chapters on the four core labor standards and acceptable conditions of work organizes these indicators into three aspects of compliance in a given country:
The first set of indicators focuses on assessment of the legal framework relating to the core labor standard (“A” indicators).
The second set of indicators addresses assessment of a government’s performance in implementation of the core labor standard (both effort and effectiveness of implementation) (“B” indicators).
The third set of indicators involves assessment of overall outcomes with regard to the core labor standard (“C” indicators).
The issues involved in identifying the indicators and in drawing inferences from them are the subject of the chapters themselves. The chapters also discuss associated factors that might either provide contextual information or serve as useful signals that there may be problems with compliance.
Both the second and third sets of indicators measure outcomes, and they are bound to overlap somewhat. The separation between the two categories is maintained, however, out of consideration that users—including the U.S. government—will sometimes want to make a judgment about a government’s performance (e.g., whether the government is “taking steps” to correct deficiencies) and will at other times want to make a judgment about overall conditions in a country.
Collecting and organizing the relevant data are only the first steps. The daunting policy challenge is to evaluate indicators of compliance and use them as a basis for policy judgments. Clearly, those judgments will always
be partly subjective—even when they are based on quantitative data—and the key question confronting policy makers who are committed to a science-based system will be how those data can best be used for fair and valid inferences. For example, in looking at freedom of association and right to bargain collectively, strike frequency and duration, union density, numbers of labor disputes, and frequency and severity of fines are important indicators (see Chapter 2). But the committee’s analysis concludes that each of these indicators could be a measure of strong compliance or, conversely, of weak compliance.
Similarly, in looking at child labor, the proportion of children who are economically active might be taken as an indicator of compliance (see Chapter 6). But the data must be interpreted carefully because an economically active child is not necessarily engaged in work that is defined as impermissible child labor in ILO conventions. In looking at acceptable conditions of work, the number of occupational injuries, both in the absolute and as a fraction of the labor force, can be an indicator of compliance. Yet, again, we note that an increase in this statistic could reflect either a larger number of injuries or improved reporting.
In each case, the context in which the actions are occurring and the specific definitions determining what is included or excluded are key. It is essential that assessors understand the source of the data and how they were collected and that they consider all of the indicators suggested for each standard, as well as the associated factors and any other contextual information that is available, when interpreting the meaning of the indicators in the database. To help assessors perform this due diligence, WebMILS provides links to the committee’s cautions about definitions, information sources, and appropriate inferences.
USING THE DATABASE AND OTHER SOURCES OF INFORMATION
Using the Data
To assess the extent of compliance with a given labor standard for an individual country, the U.S. DOL officials—or other assessors—might begin by examining relevant labor laws and evaluations of them as summarized in the database and described in more detail in the major comprehensive, systematic, and recurrent sources to which WebMILS links. As indicated in the chapters on individual standards, such sources include the
ILO supervisory mechanisms, the annual country reports, and the director-general’s global report under the Follow-Up to the ILO Declaration on Fundamental Principles and Rights at Work and the U.S. Department of State’s Country Reports on Human Rights Practices (see Chapter 2). These sources could be supplemented by in-depth individual studies when available. One example is the report to the Council of Ministers under the North American Agreement on Labor Cooperation on the effects of sudden plant closings on freedom of association.3 An assessor would need to determine whether the legal framework of a country protects that core labor standard adequately, whether it protects the core right in general but is deficient in some important aspects, or whether there are serious deficiencies in several important aspects of the core right.
U.S. DOL officials—or other assessors—would then use WebMILS and these other sources to undertake an evaluation of an individual nation’s performance—efforts and effectiveness—in complying with the core standard. In this case, the reporting of advocacy groups could also be helpful in evaluating implementation. For example, the International Confederation of Federal Trade Unions (ICFTU) and national trade union centers, the London-based Amnesty International, and other international and national human rights and worker rights groups report on specific instances of failure or broader patterns of failure by national governments in enforcing labor laws key areas. Although the latter reports are complaint driven and may be biased toward the identification of violations of worker rights (discussed in Chapter 2), they can nonetheless provide important indications of failure in implementation that could be used to cross-check other sources of information.
The assessment of an individual government’s performance in implementation needs to include both negative and positive aspects of ensuring compliance: the negative aspect is the extent to which governments leave workers alone so that they can exercise their rights; the positive aspect is the extent to which governments intervene to protect workers’ exercise of their rights. Assessment would include both proactive enforcement (inspections, reporting requirements, appropriate interventions or mechanisms to effectuate legal rights) and widely available reactive or adjudicative mechanisms for responding to individual or group claims that their rights have been
Available at http://www.naalc.org/english/publications/nalmcp.htm [October 7, 2003].
denied. The objective would be to determine the extent to which both proactive enforcement and adjudication work well, the extent to which they are weak, or the extent to which they rarely occur.
The assessment of any country’s performance is complicated because efforts and effectiveness have to be considered in the light of the resources available to the government for compliance, the poverty of the country, the number of workplaces, and the extent of other demands for government resources. In many African countries, for example, where 80 percent of the population may have no access to clean water, governments may not choose to devote scarce resources to labor standards. At the same time, it is not very costly for a government to take steps to eliminate discrimination in its own activities or to change laws that restrict the rights of workers to organize unions.
The assessment of a country’s performance also needs to include an appreciation of the extent of government support for the positive agenda of educating workers and communities about rights and remedies, building capacity among labor regulators, promoting the spread of best practices associated with voluntary compliance, and encouraging the use of independent monitoring and reporting outside the government regulatory apparatus (see International Labour Organization, 2000, Parts II and III). Some of these, such as increasing labor inspection capacity, would be costly, but others, such as publicly speaking out in favor of worker rights, would not.
In many developing countries, however, local capacity is not strong enough to pursue an effective positive agenda. In these cases, as with compliance overall, assessors should analyze the institutional and policy weaknesses affecting working conditions. That analysis may, in turn, indicate where improvements need to be made and help identify needs for financial and technical assistance to strengthen local capacities to improve labor policies. Finally, an external assessor will need to evaluate indicators of overall outcomes, keeping in mind that relevant data are often lacking or of variable quality. As noted above, the context for assessing these indicators is crucial, since high or low measures on some of these indicators can indicate either increasing or decreasing rights for workers in different situations.
In addition to assessing levels of compliance, the committee’s charge also includes the task of recommending methods to measure progress. This requires using WebMILS and the other sources identified here to look for evidence, for example, that laws have changed, more effort put into enforcement, or that market outcome indicators are improving. The U.S.
Foreign Service officers who prepare the U.S. Department of State’s annual human rights reports are instructed to note improvements, deterioration, or other changes regarding each of the rights. ILO supervisory reports also note changes in a country’s practice in comparison with earlier ILO reviews, although these are not systematically prepared for all countries. Reports from the ICFTU and other complaint-driven sources of information, such as those from nongovernmental organizations in Europe, Asia, and Latin America, may provide important insights, but they present more difficult problems in determining whether there are consistent trend lines in compliance.
In weighing the relative importance of these various indicators, context is, again, crucial. Many countries have reasonably good legal frameworks but are either unable, because of inadequate resources, or unwilling, because of political concerns, to effectively enforce labor standards. In such cases, assessors should place greater weight on the two outcomes assessments (government performance and observed outcomes) and less weight on the first assessment (legal framework). The reasoning is that the simple placement of laws on the books, without indications of effort and effectiveness in enforcement and without demonstrable results in the lives of workers, does not constitute a solid indication of compliance with the core labor standard.
Relatively greater weights for the outcome measures are not always appropriate, however. If the legal framework is clearly inconsistent with international norms, then effective government performance enforcing it should be given a negative weight, if any at all. Similarly, in judging how to evaluate trends in compliance for any given country at any time, assessors need to be alert to changes in regional or international stability—political, military, or economic—that may swamp all other determinants of working conditions from one period to the next.
The committee believes that any effort to assess the level of and direction of change in compliance should provide separate and independent treatment of each of the core labor standards and not attempt to construct a composite characterization for any given country. Favorable compliance on one standard (e.g., child labor) should not be represented as offsetting unfavorable compliance in another category (e.g., freedom of association and right to collective bargaining). Core worker rights are not substitutable: a country that eliminates forced labor but maintains discrimination in the workplace should still have the latter clearly identified.
A Framework for Analyzing Compliance Indicators
Once the information is compiled, the next step is to analyze it. In using this report and web-based information collection system, U.S. policy makers and other assessors will sometimes want to make judgments about a country’s performance with regard to all four of the core labor standards (perhaps as background for a proposed trade agreement). At other times, they will want to make judgments about a country’s performance with regard to a single standard (perhaps for use with legislation that bans the importation of the products of forced labor). Furthermore, with regard to providing assistance to help remedy any problems identified, the U.S. Congress appropriates funds to address some issues—such as child labor—separately from others, such as worker rights. These different uses affect how the information will be used.
Some experts with whom the committee consulted recommended devising a quantitative method for ranking countries according to their level of compliance with international labor standards. Such rankings of country performance in various areas have become popular in recent years. For example, Transparency International ranks more than 100 countries according to their scores on an index of “perceptions” of the level of corruption within each country. The Center for Global Development recently created an index to measure the “commitment to development” of 21 rich donor countries. The World Economic Forum has an index ranking countries according to their performance on environmental issues. The committee quickly concluded, however, that a seemingly precise, linear ranking was neither feasible nor desirable. Moreover, it also rejected the notion of using precise numerical weights for each of the indicators in order to assign countries to categories.
By contrast, when ILO experts assess compliance, they identify particular areas of a country’s law or practice that are inconsistent with international standards, they suggest ways of remedying the situation, and if appropriate, the ILO offers technical assistance. The committee believes that this disaggregated approach to assessing and promoting compliance is to be preferred and that, with the identification of indicators and creation of WebMILS, the committee has fulfilled its task.
Other members of the committee note that U.S. law requires the executive branch to assess overall compliance with these standards for a variety of purposes. In 1984, the U.S. Congress conditioned eligibility for trade preferences granted to developing countries on their performance with re-
spect to “internationally recognized worker rights.” Subsequently, these conditions were extended to a variety of preferential trade programs, as well as to eligibility for risk insurance from the Overseas Private Investment Corporation. Whatever one thinks of the desirability of these requirements, they exist and must be implemented.
Currently, however, the process for making assessments under these programs and agreements is opaque and often subject to political pressures. It is impossible to make the process purely scientific, objective, and apolitical, but some committee members believe that it could be improved by making it more open and transparent. Such openness, in turn, would make it easier for foreign governments and a variety of individuals to raise questions and contest the assessments. In addition, the DOL’s charge to the committee requires it to identify “innovative measures,” both of countries’ current compliance with international labor standards and with changes over time.
In response, the committee has developed a matrix framework as one possible way to analyze the data (see Figure 1-1). A matrix framework, with variations in levels of compliance on one axis and the direction of change on the other, might be a useful way of capturing these two dimensions simultaneously. In addition to avoiding the superficial precision implied by using specific numbers and weights to create a detailed index or ranking of countries, the matrix approach would have the advantage of giving a dynamic picture of compliance. Using this framework, assessors would carefully analyze each of the indicators in all three categories (legal framework,
government performance, overall outcomes) for a given standard and come to a judgment about the placement of each indicator along both axes.
In general, “some problems” with compliance are those that are not so frequent, broad, or severe as to seriously undermine compliance with a given standard. “Severe problems” are such as to make compliance difficult or impossible. We expect, however, that many indicators for many countries will fall into the middle category: problems that are serious enough to raise questions about compliance in important areas or for particular groups, but not across the board. For indicators in this middle category, careful scrutiny of whether compliance is improving or deteriorating will be of particular importance.
Clearly, there are other ways to design a matrix framework, and other ways to label the cells. For example, a 2 × 3 matrix that retains the three-part dynamic element but collapses the compliance decision to a binary one (in or out) is possible, but the committee believes that it would fail to reflect the complexity of the situation in most countries. As assessors work with the indicators and the matrix, they may find that four categories, such as those used by the Organisation for Economic Co-operation and Development (OECD, 2000), or perhaps subcategories, such as those used by Freedom House within the broad categories of free, partly free, or not free, are more appropriate (Freedom House, 2003). But as a starting point, the 3 × 3 matrix may be a useful way of highlighting the fact that, while some countries will fairly clearly fall into either the some or severe problems categories for some standards, most countries will fall into the middle category with respect to some indicators.
We do not attempt to delineate the thresholds along either axis that would place a country above or below the line separating the categories because we do not believe that it is possible to set thresholds that are appropriate for every country and every situation. Judgments about government performance have to reflect the resources that a government has to work with: the question is not solely whether a government is spending more or less than a given percent of its gross domestic product, but whether it is doing the best it can with what it has. And, as noted above, many of the outcome indicators can mean different things under different circumstances. The burden of demonstration, argumentation, and persuasion necessarily falls on assessors—and their critics—to defend or challenge any given assessment.
Given the necessarily subjective and contingent nature of these judgments, the major value of the matrix framework is to increase the transpar-
ency surrounding U.S. government assessments of compliance with international labor standards. It also provides a structure to reveal clearly and explicitly those areas in which value weightings or empirical inferences among assessors and their critics diverge.
Under this or any other monitoring framework, it is impossible to ensure that individual assessors are not hasty, biased, or ill-informed in making judgments about compliance. Given this reality, the matrix framework may be useful in identifying weaknesses or political biases, as well as divergences, in alternative assessments. A hypothetical example of how this might work is presented in the annex to this chapter. As illustrated there, the matrix framework exposes clearly where differences in emphasis or interpretation lie and presents a structure for open contestation among assessors.
CONCLUSIONS AND RECOMMENDATIONS
The committee concludes that the informational base for assessing compliance with international labor standards is very far from ideal. This chapter has presented examples, which can be multiplied, indicating why it is difficult for an observer to decide on the severity of a problem or its evolution.
To overcome the difficulties noted throughout the report, the committee makes the five following general recommendations:
1-1 The committee recommends that the U.S. Department of Labor improve, maintain, and update the committee’s website database.
1-2 The committee recommends that this improved website database be publicly accessible with a mechanism that allows public comment.
1-3 The committee recommends that the U.S. Department of Labor, working with other federal agencies and international institutions, support programs designed to strengthen reporting and information through capacity building in particular countries.
1-4 The committee recommends that the U.S. Department of State and the U.S. Department of Labor devote higher priority to monitoring labor standards, to developing greater expertise in this area, and to improving coordination between the two departments.
1-5 The committee recommends that the U.S. government, using agencies such as the National Science Foundation, fund research and development on methodologies for monitoring labor standards.
ANNEX: HYPOTHETICAL DEMONSTRATION OF THE MATRIX FRAMEWORK
This annex demonstrates how the matrix framework can be used to assess compliance with a single core labor standard—in this example, freedom of association and effective recognition of the right to collective bargaining—for a hypothetical country. It uses the 37 principal indicators discussed in the report: 21 “A” indicators for evaluating the extent to which a country’s system of laws and regulations are in compliance with international norms, 13 “B” indicators for evaluating the effectiveness of a government’s performance in implementing compliance, and 3 “C” indicators for overall outcomes. Figure 1-2 shows how one assessor might evaluate the legal framework in the country, government performance in implementing compliance, and overall outcomes. Figure 1-3 shows how a second assessor—or a critic of the first assessor—evaluates the same data.
What has been accomplished by utilizing the matrix framework? It leads both assessors through an extensive review of what the committee considers the major indicators of compliance. Whether each assessor has undertaken with the same degree of thoughtful integrity and brought the same degree of expertise to this task, as opposed to engaging in a mere pro forma check-off of the lists of indicators, is of course unknowable. In light of this, the format for structured comparison shows both assessors—and outside observers—where the divergences lie and invites discussion of alternative judgments and inferences.
In this case, the two hypothetical assessments differ in the importance accorded to laws that make citizenship of members a precondition for establishing a trade union (A-5) and laws that forbid sympathy strikes (A-15) (see Chapter 2). They also differ in the importance accorded to delays in the prosecution of employers who retaliate against strikers (B-2), in the importance accorded to government interference in freedom of association by declaring martial law or “state of crisis” (B-4), and in the importance accorded to frequency of labor inspections (B-8). The two assessments further differ in the interpretation of whether an increase in the number of
strikes between two periods of comparison (C-2) represents an increase in repression of workers, a relaxation in repression of workers, or more extensive reporting.
It is not possible to provide specific guidelines on where to draw the line between characterizing a country’s compliance problems as “extensive” or “severe.” But the matrix framework at least makes the value judgments and data interpretations of assessors explicit and exposes them to challenge and debate.
Elliott, K.A. (2004). Labor standards and the free trade area of the Americas. In A. Estevedeordal, D. Rodrik, A. Taylor, and A. Velasco (Eds.), FTAA and beyond: Prospects for integration in the Americas. Cambridge, MA: Harvard University Press.
Engerman, S. (2003). The history and political economy of international labor standards. In K. Basu, H. Horn, L. Raman, and J. Shapiro (Eds.), International labor standards: History, theory and policy options. Malden, MA: Blackwell.
Freedom House. (2003). Freedom house country ratings. Available: http://freedomhouse.org/ratings/index.htm [October 8, 2003].
Huberman, M. (2002, August). International labor standards and market integration before 1913: A race to the top. Paper prepared for Trinity College Conference on The Political Economy of Globalization: Can the Past Inform the Present. Dublin, Ireland.
International Labour Organization. (2000). Your voice at work: Global report under the follow-up to the ILO declaration on fundamental principles and rights at work. Geneva: International Labour Office.
International Labour Organization. (2003). Ratifications of the ILO fundamental conventions. Available: http://webfusion.ilo.org/public/db/standards/normes/appl/appl-ratif8conv.cfm?Lang=EN [October 7, 2003].
Jones, E.B. (1963). New estimates of hours of work per week and hourly earnings, 1900-1957. Review of Economics and Statistics, 45(4), 374-385.
National Research Council. (2003). Monitoring international labor standards: Quality of information, summary of a workshop. M. Hilton (Ed.), Division of Behavioral and Social Sciences and Education and Policy and Global Affairs Division. Washington, DC: The National Academies Press.
Organisation for Economic Co-operation and Development. (2000). International trade and core labour standards. Paris: Author.