So I've been doing some research on another topic I haven't really touched much as of yet. This is a topic that gets brought up frequently during the discussion about gender issues. The topic is son preference in developing countries, particularly China and India, and how that tends to benefit males. My cursory findings are that this is only half of the discussion, ignoring how a fairly sizeable proportion of the world - even if you limit yourself to the developing world - has a preference in favour of daughters.

-----

The most comprehensive study I've found on the topic is Fuse 2010, which notes that "A preference for sons is not always found; and, indeed, a preference for daughters is shown to prevail in many societies."

https://www.jstor.org/stable/pdf/26349621.pdf?refreqid=excelsior%3Acdaadcac9467d2a58832e0dafda620b6

Women were asked questions on their gender preferences, and based on their responses to these questions, the women were classified into one of four gender preference categories: no gender preference (NP), balance preference (BP), son preference (SP), and daughter preference (DP).

"NP was assigned to women who provided a non-numeric answer (e.g., “It is up to God”) as their ideal number of boys, girls, and children of either sex. NP was also assigned to women who stated that their ideal number of children of either sex is one or more, and to those who said their ideal number of both boys and girls is none. BP was assigned to women who said their ideal number of boys is equal to their ideal number of girls. Women who indicated that their ideal number of boys is greater than their ideal number of girls were placed in the SP category, and those who reported a preference for more girls than boys were placed in the DP category."

Data was examined from 50 developing countries and in total, 21.6% of women were classified as NP, 47.9% of women were classified as BP, 17.7% of women were classified as SP, and 12.8% of women were classified as DP. 12.8% vs 17.7% isn't a large difference. So it's notable, considering how much you hear about son preference and how little you hear about daughter preference, that similar percentages of women surveyed exhibited daughter preference as compared to son preference, and that balanced preferences were the most common pattern found.

The data is then broken down by region.

"Figures 1 and 2 provide visual presentations of the magnitude and patterns of gender preferences cross-nationally. Looking at Figure 1, we can see that balance is the most popular type of gender preference in all non-sub-Sahara African countries, except in Turkey and Egypt, where no specific gender is the most common of all the preferences. In half of these countries, the second most common preference is no gender preference. When we focus on the preferences for a specific gender (i.e., either a son preference or a daughter preference), we can observe a variation across regions and countries. When we compare the percentage of women who have son and daughter preferences, we find that, in all of the Latin American/Caribbean countries except for Bolivia, more women report having a daughter preference. The percentage of women exhibiting a daughter preference is especially high in the Dominican Republic (33.2%) and in Haiti (23.2). This is consistent with prior anthropological research in the Caribbean, which found that matrifocal kinship patterns are prevalent, and that daughters are thus highly valued (Quinlan 2006). Daughter preference is more common than son preference in Cambodia and the Philippines, where 26.8% and 20.9% of women have a daughter preference, compared to 14.5% and 14.5% with a son preference, respectively. In addition, the percentage of Indonesian women with a daughter preference (10.7%) is slightly higher than those with a son preference (9.5%). This is in line with the results of a study by Arnold and Kuo’s (1984), which found a tendency towards daughter preference in a few Southeast Asian countries with a Malayo-Polynesian background, including Indonesia and the Philippines. Consistent with prior studies, son preference is more common than daughter preference in the countries of Southern Asia, Western Asia, and Northern Africa. The percentage of women with a son preference is highest in Pakistan, at 35.3%; followed by Nepal, at 29.2%. Southern Asian countries also have extremely low percentages of women exhibiting a daughter preference. In Bangladesh, India, Nepal, and Pakistan, only 1.9%, 2.5%, 1.6%, and 3.0% of women reported a daughter preference, respectively."

"Turning to gender preference distributions in sub-Saharan Africa (Figure 2), we find that balance is the most common type of preference. Of the 28 sub-Saharan African countries, balance is the most popular preference in 24 countries. No gender preference is the most common type of preference in two of the 28 countries. Comparing the percentage of women with a son preference to those with a daughter preference, we find that some variation can be observed. In 16 of the 28 sub-Sahara African countries, more women have a son preference than a daughter preference. While son preference is observed in every sub-region within sub-Saharan Africa, it appears to be particularly prevalent in Western Africa. Western African countries that were, based on earlier DHS data, found to have substantial levels of son preference in the early 1990s (Arnold 1997) continued to exhibit considerable biases towards sons in the 2000s (i.e., Mali and Senegal). Today, son preference is especially prominent in Burkina Faso and Senegal, where more than 30% of women have a son preference. Interestingly, daughter preference is more prevalent than son preference in 12 of the 28 countries, and these countries are dispersed across all the regions of sub-Saharan Africa."

-----

This Slate article notes that there are "a glut of orphan boys and a paucity of girls" in Cambodia, suggesting that parents tend to more readily give up sons there. This is consistent with the daughter preference Fuse 2010 found in Cambodia. On a related note, the article also states that adoptive parents in the West tend to register preferences for girls regardless of the country they're adopting from and the sex ratio of those put up for adoption in said country.

https://web.archive.org/web/20210411002120/https://slate.com/news-and-politics/2004/01/why-do-adoptive-parents-prefer-girls.html

-----

The question will then come, "Well, has daughter preference ever had as bad effects on sons as it has had on daughters? What about the missing women?" After all, mere preference does not necessarily lead to infanticide, neglect, abortion or other such things.

The answer's yes. There's corresponding evidence of "missing men and boys", though it all largely went unnoticed when it was happening.

http://www.econrsa.org/system/files/publications/working_papers/working_paper_804.pdf

The paper notes that among black South Africans there was a predominance of girls. In fact, this pattern was found in most sub-Saharan African countries in the twentieth century. "An excess of girls was found in almost every census conducted in the twentieth century in sub-Saharan countries." The paper argues that this anomaly was caused by a combination of higher mortality of boys and a preference for girls.

Census authorities, at least as far back as the mid-twentieth century, were aware of the missing boys of Africa and realised that sub-Saharan countries during the twentieth century had a very low sex ratio of the under-five population - in fact, they were lower than those of pre-industrial and nineteenth-century Europe.

By looking at anthropological studies of sub-Saharan African tribes and recent research which uses anthropometrical data, they conclude "The higher mortality of boys could be the consequence of a cultural preference for girls."

"Studying the Herero in Botswana, Harpending and Pennington (1991) found evidence of such a preference, although the Herero themselves did not recognize this. The daughters were given better food than the sons and were more protected from risks, with a resulting lower rate of female infant mortality. This situation tends to be found in areas where females are more productive because they have a fundamental role in the workplace, and particularly where their families receive a brideprice when a daughter marries (Rosenzweig and Schultz 1982; Cronk 1991)."

Looking at the the Statistical Yearbooks of South Africa between 1955 and 1990, they note that "in all cases there was a clear excess of mortality of male children. This is consistent both with the biological explanations and with the anthropologists' findings that female children in these societies tended to be better fed so, probably, with a higher body mass index."

"Overall, studies of the twentieth century suggest that girls had an advantageous situation or, at least, were not clearly discriminated against. These results are echoed in South Africa. For South Africa, we have data on the health of children under five years of age since the last decades of the twentieth century, when improvements in nutritional status had already taken place. The results from the 1990s confirm that poor children (especially coloureds and blacks) were more likely to be stunted, and boys more than girls (Health Systems Trust 1999; Zere and McIntyre 2003; Lesiapeto et al. 2010). Poor black and coloured male children were therefore more likely to suffer problems related to poor nutrition. In the twenty-first century, in South Africa, these nutritional problems have almost disappeared (May and Timæus 2014)."

Part of the reason why daughter preference was found was the practice of brideprice in many sub-Saharan African societies. "In any case, in these societies, women are not mere family bulk but a source of family income. Interviewing Zulu parents about their daughters' brideprice, Posel and Rudwick (2014) were told it is economic compensation for the expenses involved in bringing up daughters. Some even said they wanted not just compensation but profit. Boys, on the other hand, are a drain on the budget since the family has to pay the brideprice when they marry (Anderson 2007)."

"Clark et al. (1995) detected a clear preference for girls, in the form of better care and feeding, in the late twentieth century in a rural community in Zambia. The discrimination was even more pronounced in the case of male twin births (twin killing practices). They reported a sex ratio of 92. During the twentieth century, only sporadic cases of infanticide were reported in South Africa. However, there are sources for the nineteenth century that confirm the existence of twin killing in South Africa or other infanticide practices (Badassy 2011; Devlieger 2013). They were documented among the Zulu, Thonga, Bashoto, Tswana and other groups. These practices diminished or disappeared during the last decades of the nineteenth century and the first decades of the twentieth century, continuing only in some small groups who maintained the tradition longer (Badassy 2011). Historical anthropologists who described infanticide practices in South Africa usually did not directly report gender discrimination practices (e.g. Junod 1927; Bryant 1967). However, we did find one interesting and illuminating example of gender discrimination in South Africa. Badassy (2011: 18_19) cites a case reported in The Natal Witness in 1882 of a married mother in Tarkastad (Eastern Cape) who had twins, a boy and a girl. The mother said she could not feed two babies, so she killed the boy who was the least value. This is just an anecdote, but it is clear evidence of a preference for girls amongst certain black South Africans historically."

So during the twentieth century and as far back as the nineteenth, there is evidence of neglect of boys in sub-Saharan Africa, including preferential feeding, care, and protection of girls as compared to boys, and perhaps even gender-discriminatory infanticide of boys. Interesting how THIS neglect of male children in sub-Saharan Africa went unnoticed, whereas during the latter half of the twentieth century when China introduced the one-child policy people were freaking out about son preference and sex-selective abortions of girls and female infanticide.

Regardless, this daughter preference seems to have died off in the twenty-first century with, as nutritional problems among boys have decreased over time, and sex ratios converged in the latter half of the twentieth century and continued into the twenty-first. The authors note "the socio-cultural changes experienced by South Africa, in the context of globalization, may have modified the preference for girls to the benefit of boys".

The idea that daughter preference was historically prevalent in sub-Saharan Africa in the twentieth century but has started decreasing is consistent with the findings of Fuse 2010, which finds that "In 16 of the 28 sub-Sahara African countries, more women have a son preference than a daughter preference" whereas "daughter preference is more prevalent than son preference in 12 of the 28 countries". So at the moment, about equal proportions of sub-Saharan African countries exhibit son preference as they do daughter preference.

-----

In the light of the sub-Saharan African situation and how it largely went ignored despite evidence of preferential treatment of girls, I can't help but think of the fact that the majority of data shows that boys are more likely to be undernourished than girls. I wonder how much of this is due to biology and how much of this is due to preferential treatment of girls. People don't usually study for preferential treatment of daughters, and the typical assumption that boys are weaker than girls and should be faring worse would tend to make people overlook daughter preference in regions where the higher male rates of undernourishment exists not solely because of biological male vulnerability, but also partially because males are in fact neglected and are undernourished because of it.

https://www.medrxiv.org/content/10.1101/2020.09.19.20196535v1.full-text

"76 studies reported on outcomes related to undernutrition – wasting, stunting and underweight. From this, 38/76 studies reported on wasting as an outcome with 31/38 (81%) reporting a higher prevalence of wasting in boys, 6/38 (16%) reporting a higher prevalence of wasting in girls, 1/38 (3%) reporting no difference in the prevalence of wasting between boys and girls. 68/76 studies reported on stunting as an outcome. 55/68 (81%) reported a higher prevalence of stunting in boys and 13/68 (19%) reported higher prevalence of stunting in girls. 37/76 studies reported on underweight as an outcome. 30/37 (81%) reported higher prevalence of being underweight in boys, 7/37 (19%) reported a higher prevalence of underweight in girls."

-----

While many countries in sub-Saharan Africa might have changed from having a daughter preference to a son preference, the very opposite trend can be seen in Asia, where many East Asian countries which previously exhibited son preference are changing to no preference or a preference for daughters. Japan and Korea now no longer exhibit a son preference, but actually exhibit a preference for daughters.

https://www.researchgate.net/publication/352881332_'Not_a_bowl_of_rice_but_tender_loving_care'_from_aborting_girls_to_preferring_daughters_in_South_Korea_Asian_Population_Studies_2021

"South Korea is the first country to shift from strong son preference to preferring daughters. This paper examines the factors associated with daughter preference, using data from the 2012 Korea General Social Survey, a nationally-representative survey of 1,379 people. The outcome variable was derived from the survey question, 'If you were to have one child, which one would you like to have - son, daughter, or no preference?' Multinomial logistic regression was used to examine the association between reported child gender preference and several social and cultural variables. 43 per cent of respondents preferred daughters, 36 per cent preferred sons, and 21 per cent were indifferent. The probability of preferring daughters over sons increased with exposure to Korea's social transformations (younger, more educated, and urban residents); and among those less vested in the traditional patriarchal norms (women, non-Buddhists, and less conservative views on gender roles). Other studies in South Korea find increasing intergenerational support between parents and daughters. This is no longer an agrarian society where aging parents depend financially on sons. Today people can save for retirement and have national health insurance. However, people live longer and need companionship and care which they feel daughters provide more than sons."

https://paa2006.princeton.edu/papers/60747

"Several Japanese surveys have found a trend for daughter preference by asking respondents about their gender preference. The most prominent survey is the JNFS, conducted every five years by the National Institute of Population and Social Security Research reveals that preference for daughters over sons has become increasingly common since the 1980s. The report of the 12th JNFS illustrates a dramatic shift from a tendency for son preference in the 1980s to daughter preference today (see Table 1) among married couples. For instance, in 1982, while only 48.5% of married couples whose ideal number of children was one child preferred a daughter (51.5% preferred a son), this percentage increased to 75% in the 2002 survey. Similarly, a shift from son preference to daughter preference can be seen among those who preferred to have two children and three children (National Institute of Population and Social Security Research, 2003)."

"The Study of the Japanese National Character, a large-scale national survey of adults conducted every five years, includes a series of questions concerning perception of gender differences. One of these questions asks whether they want a boy or a girl if they were to have only one child. In 1988, overall, while more respondents preferred a boy (32%) than a girl (29%), the percentage of those preferring a girl (47%) was greater than those reporting boy preference (27%) in 2003 (Institute of Statistical Mathematics, 2004). Child sex preference differs by gender, however. Among male respondents in 2003, although the percentage preferring a boy was higher (41%) than that preferring a girl (28%), the percentage gap has clearly narrowed since the 1988 survey (45% preferred a boy and 15% preferred a girl in 1988). As for female respondents, while they have tended to prefer a daughter as revealed by the 1988 study (40% preferred a daughter and 22% preferred a son), this tendency has intensified by 2003 (64% preferred a girl and only 16% preferred a boy)." 

"The Japanese General Social Survey (JGSS), a national public opinion survey of adults 20 years of age or older also asks respondents whether they want a boy or a girl if they were to have only one child. Survey results from 2000 indicated that among men, 60.5% wanted a boy and 35.2% wanted a girl, while among women, 26.3% preferred a boy and 70.2% wanted a girl child (Iwai and Sato, 2002)." 

-----

What about in history? Many societies which have been thought to primarily prefer male infants and kill female infants disproportionately have been found to be more egalitarian than previously thought. Like Ancient Rome.

https://www.livescience.com/42834-ancient-roman-infanticide.html

"The 12 babies studied in the new paper bring the total number of ancient Roman babies thought to be victims of infanticide who have undergone DNA testing to 25. Overall, there is no evidence that baby girls were killed more often."

In fact, there were laws in ancient Rome legalising the infanticide of boys, but not of girls (unless they had a deformity). The Law of the Twelve Tables notes under the section "Paternal Power":

https://avalon.law.yale.edu/ancient/twelve_tables.asp

2a. To a father ...shall be given over a son the power of life and death.

2b. If a father thrice surrenders a son for sale the son shall be free from the father.

So let's parse this. In Rome, sons could be killed. Sons could also be sold by their fathers into slavery three times and only then the son would be free. Interesting how you don't hear a peep about that.

-----

I might be making more posts on this soon, but it's interesting just how much interesting counter-narrative info I could find just from some short digging.