Devil in the details
Studies show an astounding degree of forced labour. However, for effective policymaking, these estimates must be scrutinised and contextualised
Yesterday, I got a frantic call from my friend. He works at a call centre. Completely drained and exhausted, he said that he had become a helpless slave catering to this globalised world. He complained that he was encountering forced labour even in today's 21st century. This got me thinking. Forced labour even in the 21st century is a difficult phenomenon to digest. While my friend's situation may not strictly fit into the definition of forced labour, others do face it. As per the ILO Forced Labour Convention, 1930 (No. 29), forced labour is defined as "…all work or service that is exacted from any person under the menace of any penalty and for which the said person has not offered himself voluntarily." Of all migrant workers in Malaysia's electronics industry, nearly one third are subject to forced labour conditions. Amnesty International reports that Unilever, Nestle, Procter & Gamble and six other brands were using palm oil tainted with labour abuses (including forced labour) in Indonesia. This practice is not easy to uncover, therefore, its estimation is often difficult. But, estimates suggest that in 2016, a whopping 40.3 million people were victims of modern slavery (including forced labour). These jaw-dropping figures are from the "Global Estimates of Modern Slavery" Report brought out by ILO and Walk Free Foundation in partnership with the International Organisation for Migration. Out of the 40.3 million, a majority of 24.9 million were in forced labour. While it is noteworthy that such a global estimation has been done, an analysis of both the report and its methodology reveal disturbing issues.
First, let's look at the findings in terms of the prevalence of forced labour. "Limiting the prevalence by region to forced labour…Asia and the Pacific have the highest prevalence (4.0 victims for every 1,000 people), followed by Europe and Central Asia (3.6) and Africa (2.8)." Surprisingly, one finds that the sample size for 'Asia and the Pacific' is the largest at 37,152 from the maximum number of countries – 14. This could be one of the reasons for the disproportionately high figures for this region lending a misleading picture.
Second, on the choice of sample countries, an important detail is hidden in the endnotes. "Note that while the methodology assumes the set of surveys constitutes a random sample of countries, in reality, the survey countries were selected for specific reasons. So the assumption is made but is not realised." It does not throw light on what the specific reasons might be that are leading to the arbitrariness.
Third, sample sizes are not uniform. "The sample size of most surveys covers about 1,000 persons, with the exception of Russia (2,000), Haiti (504), and some countries where multiple surveys were implemented with samples of 2,000 (or 17,000 in the case of India)." From the reports, it is clearly evident that out of the 48 countries studied, in the case of 18 countries the sample was either greater than 1000 or, in rare cases, less than 1000. For others, it was 1000. Many have pointed out the huge deviation in the sample size of only one country. India's sample size is 17,000.
Fourth, instances of modern slavery (including forced labour) have been imputed to the country of exploitation. In the case of exploitation, especially by private actors, it seems to be unfair that their nationality is not accounted for. So, it doesn't matter that foreigners exploited their domestic servants (who are residents of the country). Scholars have also argued that one should look at where exploitation happens and at whose behest. Some time back, Nike was heavily criticised for the shocking physical and mental abuse of workers making Nike's Converse shoes at an Indonesian factory. In this era of globalisation, this needs to be factored in.
Fifth, while carrying out national surveys, information was sought about the person interviewed and their immediate family. People were asked if their family members had experienced forced labour. It is surprising to note that 71,758 were self-responses while nearly 5,03,552 were proxy responses (response on family members) out of a total of 5,75,310 responses. Out of the 71,758 self- responses, only 747 were victims of forced labour at any time in the past. Out of the remaining nearly 5,03,552 proxy responses, 1,925 were victims of forced labour at any time in the past. This means that the majority of forced labour victims were identified using proxy responses. Though response errors have been accounted for, estimates would have been more credible if only personal experiences were counted.
Sixth, the refusal to answer a question was recorded in a surprising manner. It is mentioned that "such refusals were considered to be indicative of a recent experience of forced labour that the respondent did not want to reveal and discuss during the interview, perhaps out of fear of reprisal by the employer or agent. These refusals were recorded as forced labour within the last five years in the data processing of the national surveys." Interpreting a refusal in this way lacks credibility. Further, in spite of refusal, in some cases, the gender of a person and the country of exploitation has been assumed. Such assumptions seem distorting.
Finally, it is clear that these estimates intend to "inform policymaking and the implementation of target 8.7 and related SDG Targets." Target 8.7 is related to forced labour, modern slavery, human trafficking and child labour. Since the main goal is to inform policymaking, it is important that a holistic picture is presented. For this, it is necessary to understand the number of cases registered, the number of cases pending in courts, the number of people who have received relief through courts and the number of cases where forced labour could not be proven. Estimates of these would help to understand the issue better. But these issues have not been dealt with. It would not be asking for too much if one is expected to find how many of the people interviewed had filed complaints. In short, while instances of exploitation are unfortunate, un-redressed instances of exploitation might be the larger issue.
While some of these arguments may seem to be criticising small details, they are still worth pointing out since estimates have been arrived at using extrapolation. This is not to discredit the effort to bring the issue of forced labour to the forefront. Its message: "Forced labourers produced some of the food we eat and the clothes we wear, and they have cleaned the buildings in which many of us live or work," is well taken. But, as many agree, care needs to be taken in estimating it so that credible data is used to inform policymaking. The last thing that governments need is untrustworthy data, showing a misleading picture, to complicate matters and confuse policymaking. As others point out, even consumers would be influenced by these estimates affecting their choice of products. Who would want to buy a branded shoe that smells of forced labour? Credible figures would arm us in this global battle of ridding the labour force of forced labour.
(The author is a lawyer and Young Professional with Economic Advisory Council to the Prime Minister and NITI Aayog, Government of India. The views expressed are strictly personal)