International Labour Organisation’s omissions

International Labour Organisation’s omissions

Its report talks about ‘forced marriages’. Its methodology fails to delineate the concept, yielding questionable data.

modern slavery in india, slavery india, narendra modi, international labour organisation, ilo, united nations, walk free foundation, forced child labour, human trafficking, child abuse, labour laws, women harassment, Forced Marriage
Photo for Representation purpose

A few weeks ago, the International Labour Organisation (ILO) and the Walk Free Foundation published a report, in collaboration with International Organisation for Migration. The report is titled, “Global Estimates of Modern Slavery, Forced Labour and Forced Marriage”. If ILO lends its name to a report, it ought to be taken seriously. Ending modern slavery, forced labour and forced marriage are laudable objectives. The report says globally, 40 million suffer from “modern slavery”. Of them, 25 million suffer “forced labour” and 15 million “forced marriage”.

There is a riddle for small children — what is full of holes, but can hold water? As everyone knows, the answer is a sponge. A reading list of critiques of the methodology used in this report is as long as an arm. The report is full of holes. Does it still hold water? Since people have generally criticised the part devoted to forced labour, let me focus on forced marriage. Sure, there is a 1962 UN Convention on Consent to Marriage, Minimum Age for Marriage and Registration of Marriages, emphasising consent, minimum age and registration (India is not a signatory). As concepts, minimum age and child marriage are readily understood.

But this report clarifies forced marriage is not the same as child marriage. “An estimated 37 per cent of victims living in forced marriage were children at the time the marriage took place…That is, forced marriage in these estimates includes all marriages of both adults and children that were reported by the survey respondent to have been forced and without consent, regardless of the age of the respondent. Accordingly, the estimates do not include every instance of child marriage, as child marriage is not currently measured adequately at the scale or specificity required for a global estimate.” Note more than 60 per cent of the forced marriage victims were not children. Note also, the authors believe child marriage, a concept clearly understood, is inadequately measured and is incapable of yielding global estimates. However, evidently, forced marriage can yield satisfactory global estimates, despite it being much more subjective and culture specific. What is consensual, and what is not, is often difficult to determine. Therefore, take a look at the Tahirih Justice Centre’s work in the US. Most of it is on child marriages. There is a compendium on state laws against forced marriage, with the caveat, “To the Tahirih Justice Center’s knowledge, no forced marriage prosecutions have ever been brought under these laws (against a parent or anyone else).” Countries (like the UK) have legislation against forced marriages, but whether a marriage has been forced or not is decided by courts after weighing evidence.

The figure of 15 million sounds large, but it is an estimate. You take the absolute number of respondents who reported forced marriage, divide it by the sample size to obtain a rate and apply it to the population to obtain an estimate like 15 million. That’s how all estimates work, nothing wrong with that. But respondents were asked about forced marriage of self, as well as of immediate family members, and the methodology section of the report tells us family member responses are not as reliable as those of self. The absolute number of respondents who reported forced marriage was 1,415 (using which rates are worked out), of which, 271 were for self and 69 were for spouse/partner. In other words, the relatively unreliably segment is the bulk. It gets even better (or worse) than that. Read this more than once to realise what’s going on. “Refusal on any of the key questions on forced marriage, for example, the question on ‘forced to marry’ or the question on ‘did you consent to the marriage’ and the refusal to identify family members after having responded positively to at least one of the key questions. These were considered to be indicative of recent experience, or knowledge, of forced marriage that the respondent did not want to reveal and discuss during the interview. These refusals were recoded as forced marriage within the last five years in the data processing of the national surveys.” This is best described as an illustration of how responses should not be elicited.


“It is important to note that the measurement of forced marriage is at an early stage and both the scope and the methodologies are likely to be further refined. Accordingly, the current estimates should be considered to be conservative.” One can’t disagree, though I am inclined to think these estimates are confused and fuddled, rather than conservative.

It is better to stick to the child marriage issue. For India, based on the 2011 Census, there is a rather good report brought out by National Commission for Protection of Child Rights (NCPCR). It is an issue not only for girls, but also for boys. As everyone knows, there are variations between states. As everyone may not know, there are variations within states too and the NCPCR report identifies 70 districts where incidence of child marriage is high. These are spread across undivided Andhra, Arunachal, Assam, Bihar, Gujarat, Haryana, Jharkhand, Karnataka, MP, Maharashtra, Rajasthan, UP and West Bengal. There is a further filtering, based on increase/decrease between 2001 and 2011 censuses. This throws up a few surprises. Districts in Rajasthan feature in the high incidence basket (not a surprise). But except for Banswara, all register a decline between the two censuses. Unexpectedly, some districts in Maharashtra and Gujarat register an increase, such as for boys under-21 in urban Maharashtra. It is such reports one should read, not the one brought out by the ILO.