Welcome to the website for theoretical research in Anthropology, hosted in Leiden, the Netherlands. This site is devoted to the promotion and dissipation of research in cultural anthropology and related sciences, and specifically the advancement of theoretical thinking within these sciences.

 

We intend to cover a wide range of topics and express the hope that  the site  will contribute to a better understanding of humanity. Please feel free to react under the "discussion" buttons.

Most human rights organizations carry out a tremendous job in fighting race-, gender- and other forms of discrimination all over the world. Recently, however, a Dutch study (van Genugten and Svensson 2010) has indicated that a development cooperation and human rights organization, subsidized with public funds, was judged by the Dutch Equal Treatment Commission to have committed race and sex discrimination against one of its employees. The story (in Dutch) can be read on pp. 75-82 of the mentioned publication (see below for link).

 

This raises the question whether there exist more development cooperation and human rights organizations who have committed such human rights violations. If so, very little is known on the subject. How do such organizations justify their behaviour? How do employees apply their human rights? Why does that fail?

 

Those who have ideas on these topics for further study may contact theory.anthropology@gmail.com for discussion, or react on this website under "discussion". Theoretical perspectives of interest are L. Festinger's theory of cognitive dissonance, with a psychological orientation, and institutional approaches such as of M. Douglas, as mentioned by the authors of the study. Case studies might also be of interest as empirical material.

 

VAN GENUGTEN, Marieke and SVENSSON, Jörgen. 2010. Dubbel de dupe? Een studie naar de benadeling van werknemers die ongelijke behandeling aan de orde stellen. Enschede: Universiteit Twente.

Link: http://cgb.nl/webfm_send/506

Some time ago (van Dokkum 1997) I argued that Trobriand Islanders do not have one single overall belief concerning procreation but that it is likely that they have or at least had at the time of Malinowski's fieldwork conflicting opinions concerning the possibility (or not) of procreation without sexual intercourse. The debate whether Malinowski was right (or not) concerning the statement that Trobrianders literarally belief in the possibility of asexual procreation and whether this could be compared (or not) with the Virgin Birth of Jesus has become known as the "Virgin Birth" debate in the anthropological literature [1]. "Virgin Birth" is , to be sure, an unlucky term, but this terminological difficulty is not in the end decisive, because more important than the disputes about Trobrianders and about Christians are the disputes among members these repective groups themselves concerning the logical relationships they establish (or deny) between procreation and sexuality in different configurations, as indeed I argued in my (1997) article.

Below follows, in part I, the text of my reply to a critique of Mark Mosko (1998), who argued that my literal interpretation of Malinowski's literal interpretation of Trobriand beliefs in asexual procreation could not be right. Mosko (1995, 1998) seems to have based his interpretation of Trobriand beliefs particularly on an idea of Annette Weiner (1988), which I show in the reply to have a paradoxical, not to say absurd, consequence. (Note: publishing the reply is subject to reprint permission from the author, which is, of course, herewith fulfilled.)

In part II below [to follow] I will review certain more recent criticisms that have been leveled against Malinowski with respect to his literal interpretation. In one case it is even suggested that he has been fooled by his Trobriand informants. I put forward an argument in order to show that that suggestion is not credible. In fact, without Malinowski it is impossible to properly understand one specific episode in Weiner's (1988) book; something she herself, as well as most of her readers, seem to have overlooked. Malinowski error was that he believed that all Trobrianders believed in the possibility of asexual procreation. It is, however, enough to accept that some, but not all Trobrianders might have believed in that possibility. However, the fact that there is no common denominator in Trobriand beliefs on procreation seems to be one that is difficult to swallow.

Note

[1] As I understand the debate it also encompasses "necessary-but-not-sufficient" situations, in which sexual intercourse (at least one time in a woman's life) is necessary for procreation but not itself causing conception.

 

 

(I) On “Virgin Birth” and a Paradox of Procreation
 
Andre´ van Dokkum (Current Anthropology 41(3): 429-430)
Foundation for Theoretical Research in Anthropology,
Leiden, The Netherlands.

Mosko (CA 39: 685–87) criticizes my report on “virgin birth” (CA 38: 99–104), in which I argued that some Trobrianders may have believed in “virgin birth” while others may not.   Mosko focuses on the first half of this argument, stressing (p. 685) reports of a belief that

 

on the one hand, the father’s sexual actions during copulation—the repeated pounding and hammering of his penis upon the mother’s womb and/or menstrual blood (and possibly also his semen)—“feed” the fetus; on the other, the father’s sexual feeding “forms” (“molds,” “grows,” “solidifies,” “strengthens,” “stops up,” “checks,” “coagulates”) the mother’s menstrual blood to give the fetus its shape or external form.

Elsewhere (1995: 767) he cites “Weiner’s report of an indigenous distinction between the initial ‘conception’ and subsequent ‘feeding,’ ‘nurturing,’ ‘growing,’ or ‘development’ stages during ‘pregnancy’” as an important interpretation of Trobrianders’ beliefs about procreation. According to Weiner (1988:57 n. 2), “Trobrianders make [the important distinction] between conception and pregnancy, which illustrates the complementarity between the roles of women and men in procreation.” She continues (pp. 57–58, my emphasis):

 

In the first stage of pregnancy a woman conceives exclusively by an ancestral spirit from her own matrilineage. The second stage of pregnancy, that after conception, reveals the importance of patrilateral kin. Trobrianders believe that following conception, frequent sexual intercourse helps to develop the fetus.   . . . The claim to the father’s contribution is so important that it is . . . expressed as necessary to the growth of the fetus.

Weiner’s two-stages theory cannot be considered a decisive solution of the “virgin birth” debate, because it can lead to paradoxical predictions. When a husband leaves his wife around the time of conception and returns to her after the child’s birth (or late in the pregnancy), the two-stages theory, with its necessity clause, predicts that he can accuse her of adultery on the basis of the growth of the child. The paradox is that this may occur even when she has been faithful to him and the child can only be considered his own.

 

Malinowski (1954[1916]: 224) writes:

 

When I asked my informants what would happen if a woman became pregnant in her husband’s absence, they calmly agreed that such cases might occur, and that there would be no trouble at all. One of them . . . volunteered his own case as an instance in point. He went to Samarai [New Guinea] . . . and stayed there for a year, as he said, during which time his wife became pregnant and gave birth to a child. He returned from Samarai, found the child, and it was all right. On further questioning, I came to the conclusion that the man was absent for about 8–10 months, so there is no urgent necessity to doubt the virtue of his wife, but it is characteristic that the husband had not the slightest tendency to count the moons of his absence.

The relative time order in the informant’s account is that the first stage of pregnancy occurs after his departure, and this is something the two-stages theory can handle. But then the second stage falls completely into the time of the man’s absence, and here Weiner’s theory does produce an accusation of adultery. It should not do this, for the accusation could well be false. In any case, the informant does not consider the two-stages theory relevant.

 

This counterexample holds also for Mosko’s own argument. Since the man believed that all of his wife’s pregnancy occurred after he left, he had no reason to believe that a hammering penis had anything to do with the feeding of the fetus. It seems that the second stage is not of decisive importance.Weiner’s two-stages theory consequently reduces to the statement that women “conceive exclusively by spirits,” and we are back at the point where the whole “virgin birth” debate started.

 

In order to explain the failure of the two-stages theory we must look at its remarkable temporal set-up—first conception, then intercourse. However, those who reject the belief in “virgin birth” will be more likely to believe in a reverse time order—one in which the second rather than the first stage of pregnancy can do without intercourse. The implicit assumption behindWeiner’s theory is that husbands stay with their wives during the second stage, but this is not always the case. It may be possible that some Trobrianders have stated the two-stages theory, but this should be taken with due observance of the implicit assumption.

 

This shows again how difficult it is to summarize Trobriand beliefs on procreation in a single overall theory. Powell’s (1968:652) suggestion that a “combination of the opposed academic hypotheses might explain the total data recorded better than either one of them” has received less attention than it deserves. The best explanation of the controversial nature of the debate on Trobriand “virgin birth” is that the culture as a whole is itself inconclusive on the topic. “Virgin birth” cannot be considered a Western ethnocentric notion (Mosko, p. 687), because Trobrianders themselves dispute about it in their own intracultural discussions. Weiner’s court trial example (1976: 122; 1988:55), in which the two-stages theory is not mentioned, is an illustration of this.
References Cited
Malinowski, Bronislaw. 1954 (1916). “Baloma: The spirits of the dead in the Trobriand Islands,” in Magic, science and religion and other essays, pp. 149–274. New York: Doubleday.
Mosko, Mark S. 1995. Rethinking Trobriand chieftainship. Journal of the Royal Anthropological Institute, n.s., 1: 763–85.
———. 1998. On “virgin birth,” comparability, and anthropological method. Current Anthropology 39: 685–87.
Powell, H.A . 1968. Correspondence: Virgin birth. Man, (n.s.), 3:651–53.
van Dokkum, Andre´. 1997. Belief systems about virgin birth: Structure and mutual comparability. Current Anthropology 38:99–104.
Weiner, Annette B. 1976. Women of value, men of renown: New perspectives in Trobriand exchange. Austin: University of Texas Press.
———. 1988. The Trobrianders of Papua New Guinea. New York: Holt, Rinehart and Winston.

(II) Comments on Malinowski's findings about Trobriand beliefs on procreation

 

(text under construction)

 

Mosko (1995: 767)   mentions especially Weiner's writings about Trobriand beliefs on procreation in her (1988) as decisive. I will therefore deal with the problems of Weiner's analysis first.

 

As indicated above, Weiner (1988) writes that Trobrianders believe in two stages of pregnancy: (1) conception; and (2) subsequent development of the fetus. The first stage, conception, would occur “exclusively by a ancestral spirit” (1988: 57; my emphasis; see also p. 55); the second would involve the sexual contribution of a man to the growth of the fetus. Weiner states (1988: 58), that Trobrianders believe that the contribution of a male is necessary for the second stage.

 

It is unlikely that all Trobrianders would have agreed that sexual intercourse is necessary for the growth of the fetus. (On p. 58, Weiner does not explicitly mention intercourse in relation to this necessity, but considering p. 57 it is difficult to see that she is not referring to intercourse. If Weiner means on p. 58 something else than intercourse, then the reference to intercourse on p. 57 is without explicit clause of necessity, which is my argument above and makes Weiner's exposition without sense. It is probable that she referred to intercourse.) If they strictly and universally believe(d) it, it would forbid husbands to travel during their wives’ pregnancies, and this does not seem to be the case, since it occurs every once in a while that husbands travel during their wives’ pregnancies.

The study mentioned below about child labour in India is highly recommended. It provides empirical material to support the theory that a rise in the education levels of the poor helps in tackling the dual problem of child labour and poverty. Special attention is given to the girl child.

 

The study is written by Shobha Raghuram and Puja Jain and is available on the Indian National Commission for Protection of Child Rights website. Being an Indian study, it provides a powerful counterexample to cultural-relativistic arguments with respect to child labour, about which below.

 

(NCPCR Policy Dialogue Series),

National Commission for the Protection of Child Rights, India, 2008.

 

Cultural relativistic arguments are at times put forward when it is argued that applying the international ILO-norms to child labour in developing countries would be a specifically Western (or Euro-American) idea that might not be applicable to situations in developing countries. Such arguments go more or less like this:
"(1) It is known that in the West there is no or negligable child labour and in (e.g.) India there is.
"(2) These differences are to be explained by the fact that Europe and India have different cultures.
"(3) From the viewpoint of Western culture one cannot criticize Indian culture.
"(4) Therefore Westerners should not say that India must ban child labour."

 

There are at least two flaws in the argument. The first is that cultures would be homogeneous; the argument ignores the fact that within cultures there can be debates on child labour. The second is that in the West as well the ban on child labour from the nineteenth century onwards was not obvious. Child labour existed and so one would have to argue that it was Western culture to have child labour. But people disagreed intra-culturally about what was the best way to go: some found existing legislation against child labour not strict enough, others saw not much harm in labour for children. Reference to "culture" does not give us a definitive clue as to what is an appropriate moral reaction to the phenomenon of child labour. Such a reaction involves what people think about some existing culture. In fact, debates in nineteenth-century Netherlands were strikingly similar to those we see today - see Schenkeveld (2003) - including the arguments, pro or con, that povery causes child labour.

 

A concrete example of such a reasoning may be found in the following quotations (Pierik and Houwerzijl 2002):

    "For one thing, the socioeconomic situation in Bangladesh or India is very different from the socioeconomic situation in the USA or the Netherlands. Moreover, the fact of cultural difference makes it impossible – or at least debatable – to simply translate liberal norms and values to non-Western societies" (p. 3). "We should not start from Western policies aimed at children growing up in Western societies because their situation might differ fundamentally from children growing up in developing countries. Instead we should seek the core values underneath the Western policies towards children and seek to apply them in our policies against child labor abroad" (p. 4).

 

Pierik and Houwerzijl acknowledge that the Western policies themselves have been changeable (pp. 5-8), but then go on to write:

    "Although most conventions on child labor are dominated by Western societies, precisely these societies have a very a-typical idea on the role of work in childhood. It is therefore important to debunk the 'Western' attitude of the ILO towards childhood as laid down in its Conventions on child labor. As William Myers [note omitted] argues: 'Sensitivity to culture and the way it mediates the effects of experience on children by no means invalidate the idea of global child labor standards but do suggest the need to carefully distinguish between values based on a broad representation of human experience and those that are merely Euro-American ethnocentrism projected by economic and political power onto a global arena'" (p. 14).

 

This line of reasoning ignores that people in developing countries could also identify with the ILO norms and consider these as their own. Classifying these norm as merely 'Western' is simply unempirical, as is testified by Raghuram and Jain's report (and many other publications produced by scholars in developing countries). Not to recognize the efforts of scientists to promote ILO-norms against child labour in developing countries themselves (as well as the efforts of activists, teachers and government officials), is a formidable lack of sensitivity that should be thoroughly addressed.

 

© André van Dokkum 2009

 

References:

PIERIK, Roland, and Mijke S. HOUWERZIJL. 2002. Western Policies Towards Child Labor Abroad. Tilburg: Tilburg University. Third draft, published on:

http://www.tilburguniversity.nl/globus/publications/childlabor_Roland_Pierik.pdf, accessed 5 april 2009.

SCHENKEVELD, Willemien. 2003. Het kinderwetje van Van Houten. Sociale wetgeving in de negentiende eeuw. Hilversum: Verloren.

This book presents new empirically based and theoretically informed studies on the contemporary social and economic dynamics of Africa, dealing with developments in the arenas of politics, economics and cultural struggle. These domains are closely interlinked. In their widest definition, culture and politics intermingle and recombine in surprising and sometimes disturbing ways. They always have a definite economic logic as well, informing value commitments and behaviour in the broader sense. Politics and economic life in Africa have, perhaps more visibly than elsewhere, influential and cultural aspects and referents, such as religion and ethnicity, which often play a constitutive role. 'Culture' and its symbolism are used instrumentally in the political, economic and social struggles in today's Africa, marked by a preoccupation with 'development'. The studies in this book underline the interplay of new hegemonic struggles of a material but also ideological nature.

 

Editors: J. Abbink and A. van Dokkum

Free download at: https://openaccess.leidenuniv.nl/dspace/handle/1887/13060

For a paper copy, please see the African Studies Centre's web page:

 

Gradually more and more publications appear in which a post-postmodern era is announced. One of these is Iris van der Tuin's "Jumping Generations. On second- and third-wave feminist epistemology" (2009), which specifically deals with generations of feminist epistemology. Van der Tuin makes several highly valuable points that can be stimulating for cultural anthropology, considering the epistemological uneasiness generated by the debates in the Writing Culture volume (Clifford and Marcus 1986).

 

It is common to distinguish between "first", "second" and "third" waves of feminism in the activist and perhaps post-activist) field. A different tripartition with a shorter time span was coined by Sandra Harding in the field of epistemology: feminist "empiricism", "standpoint theory" and "postmodernism" (see Anderson 2009[2000], for a profound overview of these forms of epistemology). Feminist empiricism acknowledges Quine's insight of theory-laden observation; male "bias" would induce limitations in the proper handling of scientific observations, but put positively this also means that feminist "bias" can help in producing observations that would otherwise been impossible or escaped our attention. Feminist standpoint theory holds that certain groups in society have "privileged access" to specific types of knowledge (talking mostly of social sciences), usually referring to representatives of subaltern social categories. This allows standpoint theory to analyse society from the angle of intersected social categories, for example "black women".  Feminist postmodernism, then, professes that such analyses of social categories are "essentialist" and "discursively constructed" and thus tries to "deconstruct" these, stressing the fluctuating character of meanings, including that of concepts concerning gender (and other) differences.

 

Specifically the third wave and postmodernism seem to coincide. Van der Tuin observes that later "waves" tend to have the pretension to overcome the previous one(s), declaring earlier  achievements is fulfilled and therefore no longer relevant. Generations of academic feminisms then become sharply distinguished. Van der Tuin proposes an alternative to such thinking, drawing an analogy from biology. Just as genes can be transposed ("jump") across locations in the genetic material, so can ideas from one generation be transposed on another (sometimes also from the younger to the older). "Jumping generations" is a methodology with which both linear conceptualizations of time and space, and the trap of non-exhaustive dichotomies per se, can be avoided" (van der Tuin 2009: 25).

 

This upsetting of the idea of clearly distinguishable generations has two consequences:  In the first place, it views generations as non-dialectical with respect to each other; in ihe second place the method and its object are, although not identical, also not separate, so the method and its object are partly conflated (van der Tuin 2009: 18). This must be so: a non-reflexive  generational theory would be capable of being separated from other generations  and thus violate its own conception of "generation". In other words: the theory should be non-foundational. Materialism, at last, brings in the recognition that matter may be relevant in the analyses, non-dialectically different from the postmodern stance that "there is no outside-text" (Derrida).

 

How can this be understood outside van der Tuin's article? One answer to this question within anthropology lies, in my view, in the way how postmodernism treated Lévi-Straussian-style structuralism. In its original form, such structuralism stressed the mathematical structuring of human society and mental expressions such as myth. Because the chain of explanation had to end in binary opposition, Lévi-Strauss had to assume that these binary operations were unconsciously incorporated within the human mind. In this sense, the theory was thoroughly foundationalist. Interestingly, Lévi-Strauss revokes this idea of unconsciousness (and thus in this sense admitting reflexivity of the mind) in his discussion with Sartre in The Savage Mind (1966[1962]). Postmodernist criticism, however, focussed on the initial structuralist formulations and "deconstructed" structuralist analyses by stressing (not so much poststructuralist but rather ultra-structuralist) that elements within a symbol system refer only to each other in a never-ending sequence of internal reference and therefore cannot provide a conclusive understanding of some reality outside the referential system. Projected onto anthropology as a discipline this meant that anthropologists could at most only understand the theories they are setting up but not societies through the theories. Hence the "crisis of representation".

 

However, had another stance towards foundationalism been taken, a different epistemological situation would have emerged. For it was not Lévi-Strauss mistake that sometimes social phenomena can take structuralist forms. His omission was not to incorporate possibilities for non-foundationalist (or sometimes "anti-foundationalist") structures in the formal models he proposed, although he can hardly be held accountable for that since anti-foundationalist modelling was not yet well-known.

 

I am referring to anti-foundationalism as described by Aczel (1988) and Smith  (1996). For example, if we define the sets A = {A, B, C, D, E} and B = {A, B, C, F, G}, these sets contain themselves as well as each other, but they have also each unique elements. This is not expressible in classical Lévi-Straussan structuralism (nor in standard mathematics). However, although one could maintain that the "containment sequence" A - B - A - B - etc. is a sort of Derridan "infinite différance", having no end but also no beginning, it is undeniable that sets A and B are finite entities. They also contain matter that has no further deferral capacity. Such an approach thus combines non-dialectical distinction with identification.

 

In a similar vain, van der Tuin shows new insights into how a post-postmodern epistemology could be achieved. Feminist epistemology, of all ages, will be thereby be of continuing inspiration. For anthropology, it is neither a separate nor totalizing way of looking at societies, but a valid source for construing ways of using one's eyes.

 

As an example, the topic of female genital cutting may be mentioned. Probably anthropological research on this type of practices is much inspired by normative ideas and thus may stand as an example of values and theories influencing empirical research about the practices. On the other hand, those who underwent or ordered (or, indeed, both) cutting have the same factual understanding of it as researchers, untying the link between fact and theory again. One might try to look for meaning from the perspective of the people who order cutting, using a socio-cultural framework. This cannot provide a complete understanding either, since not all cultural insiders may agree with the practices. Postmodernism would stress the differences within socio-cultural environments, but its stress on the infinite deferral of meaning is also unsatisfactory because the stakes are quite clear: some want to abolish and others to continue cutting. leaving no "infinity". Feminist epistemology can help here by non-dialectically reviving  the creativity of its earlier generations and applying it reflexively where it is needed. Should a "cultural imperialist" version of "Westerners" demanding the abolition of cutting be avoided, the external support of those activists trying to change their societies from within could be promoted. For non-foundationalism, this is entirely within its range of applications.

Bibliography

 

Aczel, Peter. 1988. Non-well-founded Sets. CSLI Lecture Notes No. 14. Stanford: CSLI.

http://sul-derivatives.stanford.edu/derivative?CSNID=00000056&mediaType=application/pdf

Anderson, Elizabeth. 2009 [2000]. Feminist Epistemology and Philosophy of Science, Stanford Encyclopedia of Philosophy.

Clifford, J. & G.E. Marcus (eds.). 1986. Writing culture: The poetics and politics of ethnography. Berkeley: University of California Press.

Lévi-Strauss, Claude. 1966 [1962]. The Savage Mind. Chicago: Chicago University Press.

Smith, B. Sidney. 1996. Hypersets. Cambridge: University of Cambridge.
http://bsmith7.asp.radford.edu/Hypersets.pdf

van der Tuin, Iris. 2006. "Jumping Generations" (Krisis dossier), Krisis, 7(3): 28-32. http://www.krisis.eu/content/2006-3/2006-3-04-vandertuin.pdf

Duprez and Baland (2004) study some economic effects of labeling programs with respect to products certified with respect to the issue of child labor. Their analysis of social labeling in paragraph 2.2 is straightforward. Assumed is a situation in which a Southern country produces products that may only be distinguished by their use (or not) of child labor in manufacturing them. These products are then consumed by a Northern country and by the Southern country itself, according to price mechanisms combined with a parameter indicating disutility of unlabeled products. In the concerned model, consumed products are individually certified when manufactured without child labor, and applying the model this results in a shift of adult workers from the manufacture of unlabeled products to labeled ones. When the Northern demand for labeled products is not enough to exhaust the Southern production capacity, Southern demand will shift according to the established price mechanisms to unlabeled products and there will be no change in the incidence of Southern child labor. This might be an explanation why labeling seems to have little or no effect on child labor in empirical situations.

In paragraph 2.3 and 2.4 of their paper, Duprez and Baland also present a model of geographical labeling, in which two Southern countries A and B produce identical products. The products of either of the two countries (say A) are labeled when that country’s incidence of child labor is smaller than that of the other (B) for identical prices of the produced goods. In fact, the two Southern countries are identical except for their child labor figures. As in the model of social labeling, price mechanisms (in combination with a disutility parameter) determine consumption and production. The authors then claim in proposition 3 (p. 12) that application of their model may actually result in an increase in total “worldwide” child labor when the label would be effective in lowering child labor incidence in country A. Because of the price mechanisms, the increase of child labor in country B would be larger than the decrease of child labor in country A. Since we have here a counterintuitive and, if true, unwanted result much stronger than the neutral result we had with social labeling, it is worthwhile to look at the arguments leading to proposition 3.

Child labor is considered to be linearly dependent on the quantity of goods produced (p. 5), where the production (like demand) in turn is dependent on prices through price equilibrium mechanisms. We may therefore establish functions of the incidence of child labor depending on prices, lA(p) and lB(p) > lA(p) for any p, of which the derivatives l’A(p) and l’B(p) are identical and given as l’(p) (p. 10). I will study here the situation l’(p) < 0 only (in the authors’ argument, l’(p) > 0 would lead to a decrease in worldwide child labor, while increasing in country A, a situation we can safely put aside for the present discussion).

The starting point of the model with two Southern countries then is that there is only trade between the Northern country and the each of the Southern countries and no mutual trade between the Southern countries. Each Southern country, however, does consume its own products according to identical price mechanisms. Before introducing the geographical label, the price p is set equal for countries A and B as deriving from the total supply and demand (in North and South) with respect to the Southern goods before
introducing a label. Country B will have a surplus supply relative to A because B is assumed to put more children to work and thus produces more goods. After introducing a label certifying country A as a country that has less child labor than B, country A sells more goods to the Northern country, while B no longer sells goods to the North. The label helps to decrease child labor in A, but as the price for unlabeled goods decreases, child labor in B increases (see p. 11; proposition 2). Proposition 3 now states that the decrease of child labor in country A will be less than the increase in B, so that the end result would actually be a total increase in the incidence of child labor.

Precisely, proposition 3 states “If l’(p) < 0 […], the net effect of a marginal geographical label is a raise […] in worldwide child labor.” This raise, however, is not supported by the argument following proposition 3. Such an adverse net effect of a “marginal” label can only be considered proved for all conceivable situations if the sum of all marginal changes can be straightforwardly determined. This is not possible with the general set-up the paper gives concerning the functions lA(p) and lB(p). Judging from figure 1 on p. 13 of the paper, it seems that these functions are perceived to be linear, but this is defined nowhere in the course running to proposition 3 (it is only defined that the quantity of goods produced is linearly dependent on labor input, but this in itself says nothing about price dependency). While l’A(p) = l’B(p) = l’(p), l’(p) does not have to be constant for all p, and lA(p) and lB(p) may be curved. If we then consider that lA(p) and lB(p) may be more elastic for higher p, it is not necessarily the case that the increase in lB(p) outweighs the
decrease in lA(p).

For a quick understanding of this, consider that the production functions in the authors’figure 1 may run somewhat more horizontal above the “p” line than Duprez and Baland give them now, while beneath the “p” line these lines may run more vertically than given. The shift of “xA(pA)” relative to “xA(p)” may be more than the shift of “xB(pB)” relative to “xB(p).” [note] This means that proposition 3 not only is insufficiently proved, it is also possible to actually construct counterexamples to it, so that labeling does decrease the worldwide incidence of child labor, even if it increases in country B due to price
mechanisms.

It is not unreasonable to assume that the incidence of child labor is less elastic at lower prices. Because country B has already more child labor than A from the start, we can assume that country B experiences more disutility from putting children to work than A, as specified in the utility functions V on p. 5 (note the limit behavior of V). While country B would experience an increased demand for child labor as a result of an increased demand of the goods produced, the increased demand for child labor would increase disutility at the production side at an already relatively depressed utility function value.

On p. 10 Duprez and Baland indicate that labeling may result in very diverse outcomes: “If countries were distinct one from another in other respects [than child labor incidence] almost every result could be obtained.” It is the question, then, why labeling itself would have such a straightforward effect as given in proposition 3. I have shown in this commentary that the application of Duprez and Baland’s geographical model can just as well lead to ambiguous results. This puts into doubt the perceived role of price mechanisms as Duprez and Baland use them. The reader may recall that it was assumed that country B had already more child labor than A at the same price level, other things equal. So the assumption in fact denies the explanatory power that is indispensable in the sequel of the argument. This is not very convincing. In my view, a more promising way to investigate the incidence of child labor would be to focus on the utility functions in which the incidence of child labor is incorporated. Apparently the valuation as to the (dis)utility of child labor may differ from country to country and this could conceivably be a more important determinant of the incidence of child labor than price mechanisms.

At least there is no reason at all to accept the statement that labeling would be capable of increasing child labor. Adding to that that labels fulfill a second-order role with respect to consumer awareness, i.e. not only in applying utility functions but also in influencing these, their usefulness, I think, still stands.

 

REFERENCE
DUPREZ, CÉDRIC, AND JEAN-MARIE BALAND 2004. ‘“Made in dignity”. The effects of labelling on child labor.”
http://www.eudnet.net/Member/afd_2004/Duprez_and_Baland.pdf

or see:

(functioning of these links cannot be guaranteed by the present website).

 

Note:

To be more precise concerning the construction of a counterexample, putting the slope of the two production curves in figure 1 more horizontal above the “p” line should not be done so drastically that the combined production curve would become lower than the combined demand curve leading to the equilibrium price “p.” If readers now imagine, for the sake of easiness, straight curve segments starting slightly above and slightly below the “p” line, with slopes and positioning relative to “p” identical for both countries, they may further imagine that for a very small neighborhood around the “p” line, we can always find a second degree function connecting with the two straight segments for each production curve, so that the three parts of each curve together form a continuous and differentiable function on the total relevant p domain. The two resulting curves satisfy the constraints given by Duprez and Baland, including the requirement that l’A(p) = l’B(p). There may be a slight change in the level of “p” relative to the Duprez and Baland value, but this change can be considered insignificant when put against a large change in specifically the “xB(pB)” shift. In fact, putting the bend under the “p” line and only making the lower segment more vertical may be enough to obtain the result that the change in country B is less than in A.

 

© A. van Dokkum