Most if not all cultures have gendered divisions of labor, but the specifics of what counts as “women’s work” versus “men’s work” vary considerably from one culture to another. There are some broad generalizations that apply pretty widely cross-culturally; hunting tends to be a masculine task, for instance, and childcare a feminine one. When it comes to more complicated tasks of more recent origin, however, the gender designations can seem pretty arbitrary. This is particularly obvious when it comes to agriculture, which developed several times in different parts of the world with different crops. In many societies tending crops is a feminine task, perhaps because it is similar to gathering wild plants, which is often a responsibility of women. This was historically the case, for example, among the indigenous peoples of the Eastern Woodlands of North America, where women tended the fields while men were responsible for hunting, war, and diplomacy. In other parts of the world, on the other hand, such as most of Europe, agriculture was traditionally considered a masculine pursuit, and men tended the fields while women stayed at home and took care of cooking, cleaning, and childcare.
Why would this be? Is it mere chance that agriculture developed different gender associations in different societies, or is there some pattern? A recent draft paper by Alberto Alesina, Paola Giuliano, and Nathan Nunn takes a look at one hypothesis, originally advanced by Ester Boserup, to explain cross-cultural differences in the gendering of agriculture. Boserup proposed that cultures in which farming is done primarily by men tend to farm with plows, while those in which agriculture is done by women use other agricultural techniques. Furthermore, she argued that these two types of agrarian societies tend to differ systematically in other ways as well, particularly with respect to gender roles. In plow societies women tend to stay at home and tend to household tasks while men are out working in the fields, and in many cases they develop highly elaborated systems of gender role differentiation with men in a clearly dominant role. This has historically been the case especially in the Near East and most of Europe, as well as in other areas such as northern India. In places without plow agriculture, however, societies tend to have less rigid gender role definition and more flexibility in acceptable economic activity for women. This is the case in most of Africa, the Americas, and southern India. Strikingly, these differences in economic role for men and women in plow societies seem to persist even when societies industrialize: men take the manufacturing jobs outside the home instead of working in the fields, but women still stay at home rather than working. Very recently this has begun to change, especially in the wealthiest societies, but there is some evidence that the pattern has been surprisingly persistent.
The proposed mechanism behind this is that plow agriculture, unlike other techniques, requires considerable physical strength to push the plow, which makes it more suited for men, who are on average physically stronger than women. Once they have gained a predominant role in agriculture, which is the mainstay of the subsistence system, men go on to become dominant in other parts of society as well, and cultural norms begin to reinforce male superiority on an ideological as well as an economic level.
An interesting idea, to be sure, but is it true? This is what Alesina, Giuliano, and Nunn set out to determine, using elaborate statistical techniques. This paper is an example of an increasing trend among some economists of analyzing anthropological questions using the sophisticated quantitative techniques that are standard in economics but vanishingly rare in anthropology. Archaeologists do use statistical techniques quite a bit these days, but the stuff in this paper blows away anything I’ve seen in archaeology, and sociocultural anthropologists hardly seem to use statistics at all. This paper is an example of what serious quantitative social-science research looks like as applied to the subject matter usual addressed by anthropologists, and the difference between these techniques and those used by anthropologists shows why anthropology is not a science. That’s not to say that these techniques are necessarily better than those used by anthropologists. In fact, I think there are a lot of drawbacks to this kind of research. Regardless of whether this kind of research is useful or not, however, I just don’t see any way to avoid the conclusion that it’s fundamentally different from what anthropologists do.
Be that as it may, this is a very impressive paper. The authors start from the assumption that plow agriculture works better for some crops than for others. Plows are expensive, as are the draft animals necessary to pull them, so societies are unlikely to adopt the plow if they can grow enough food on the land available to them more easily. This will be the case if the land is suitable for growing crops that don’t need plowed fields to grow effectively. These include tree crops, roots, and tubers, as well as some grains such as corn, millet, and sorghum. Other grain crops, such as wheat, rye, and barley, require extensive preparation of the land by plowing before they will grow effectively. The idea, then, is to identify the parts of the world suited for “plow-positive” and “plow-negative” crops, then to identify the traditional agrarian societies that have historically inhabited these areas and to check that their agricultural techniques corresponded to the crops to which their land was suitable. Using several databases of ethnographic documentation and soil suitability data, the authors confirmed that this in fact the case. They then used some fairly ingenious analysis with geographical information systems to calculate percentages of the population whose ancestral ethnic groups used the plow in countries around the world and compared these percentages to replies to worldwide surveys of attitudes toward gender roles and statistics on women’s participation in the workforce.
The results confirmed Boserup’s prediction: countries with higher percentages of people from plow cultures had lower numbers of women working outside the home and gender attitudes among both men and women more focused on male superiority. To make sure these results were really associated with plow use rather than something else they controlled for a wide variety of potential confounding variables, and the results were still quite robust. They also did a separate analysis of workforce participation among female immigrants and daughters of immigrants to the US, to control for the possibility that institutional structures in countries with histories of plow use rather than cultural factors were causing the effects on women’s workforce participation there. The results remained the same, with female immigrants from plow cultures less likely to work than other immigrant women and the same effect for daughters of immigrants, with the effect strongest when both parents were from the same ethnic background.
These results aren’t definitive, of course. The paper is certainly pretty rough, although more in the writeup than the data. Also, with any statistical analysis like this there is always the possibility of additional factors that the authors missed that could be causing these effects. In this case, however, the authors did a very good job of considering and controlling for a wide variety of possible confounding factors. This paper is a good example of how interdisciplinary research can work well. In this case the data and many of the crucial assumptions behind the analysis came from anthropology, while the analytical tools came from economics. This kind of research won’t replace traditional anthropological research, but it is a useful supplement to it. Multiple lines of evidence are always good when evaluating complicated questions such as those that arise in studying human societies.
To bring this back to the usual subject matter of this blog, another possible factor in whether or not a society uses the plow is of course the presence or absence of draft animals. The general lack of draft animals in the Americas was probably one contributing factor to the lack of plow agriculture. We’ve seen a similar explanation advanced for the lack of wheeled vehicles in the Americas as well. However, given that maize and other indigenous cultivated plants like potatoes and manioc don’t require use of the plow, it’s likely that the crop-based explanation advanced by this paper is more important in explaining the non-plow nature of American agriculture. Regardless of why, however, it is certainly the case that the Americas did not develop plow cultures like those of the Old World, with their extreme gender inequality and ideology that a woman’s place is in the home.