Why do women in South India have more freedom than their northern sisters?

Everyone knows that South India and North India are very different in culture, language, and socio-economic development. But the most dramatic regional disparity may be in gender relations.

Southern and North Eastern women are more likely to survive infancy, be educated, marry later, choose their own husbands, interact more closely with their husbands, bear fewer children, own more assets, exercise more control over their dowry, socialise with friends, move more freely in their communities and work alongside men.

In North and North-West India, women are much more constrained and sex ratios are far higher.

Education, paid work and age are all associated with greater economic and physical autonomy. But even if a woman completes secondary school, she is less likely to choose her husband if she lives in the North.

Region is a strong predictor of female survival, literacy, autonomy, employment and independent mobility. A woman with the exact same household wealth, caste and religion will likely have more autonomy if she lives in the South.

These regional trends do not hold for all aspects of gender, however. Female political representation and independent property ownership are low nationwide.

Deep roots

In 1900, girls were more likely to survive infancy, go to school and marry later if they lived in South or North East. Going further back to 1800, 90% of recorded sati cases occurred in Bengal, with far fewer in Madras and Bombay.

When Madras was ravaged by famine in 1876-’78, sex ratios remained even. But in Punjab’s famine of 1896-’97, little girls died disproportionately.

In 1880, girls in Kerala and Karnataka married at 15 or 16. Rajasthan took another century to catch up.

Educated women in 19th-century Maharashtra ran their own journals, to discuss women’s lives and social reforms. Rather than passively accept employment discrimination, they organised. When Marathi women travelled to Calcutta they were struck by marked differences in gender relations. “A woman can scarcely stand in the presence of her relatives, much less before her husband,” remarked Anandibai Joshi, staying in Calcutta. “Her face is always veiled. She is not allowed to speak to any man, much less laugh with him.”

Whenever she went to the bazaar alone, Anandibai was pelted with stones. Many Bengalis bitterly opposed the Age of Consent Bill (1891), proposing to raise girls’ age of consent to 12.

India’s gender divergence

This blog reviews the existing literature on poverty, colonialism, matriliny, cousin marriage, conquests and purdah, labour-intensive cultivation and ancestral crop yields.

Poverty in North India

Poverty can thwart progress towards gender equality. Poor girls usually quit school early, bear many children, become burdened with care-giving, then struggle to accumulate the capital, knowledge and networks to challenge dominant men.

So, does poverty explain India’s gender divergence?

Well, some Northern states – like Uttar Pradesh and Bihar – are very poor.

But expenditure is also low among North Eastern hill tribes, and yet women maintain a relatively high status, move freely in their communities, and have long been integral to shifting cultivation.

Haryana and Punjab meanwhile are two of the richest states in India with the worst child sex ratios: 830 and 846 girls per 1,000 boys. Sex ratios are worsening in the north-west, alongside economic growth.

Regardless of household income, a woman is less likely to have been to school if she lives in the North (ie Jammu and Kashmir, Himachal Pradesh, Haryana, Punjab, Uttarakhand, Rajasthan and Uttar Pradesh). Nowadays the North does not perform badly on literacy. But the gender gap in education is 26% in the North, 9% in the South.

Female education is improving in the North. We might expect more skilled women to gain economic autonomy, expand their networks, broaden their horizons, demonstrate their equal competence in socially valued domains, support elderly parents and become valued as providers. That is certainly what happened in patrilineal China and Taiwan, but not in India.

Regardless of their qualifications, rural women tend to retreat from the labour force when their families are economically secure, especially in the North. Rural women gain status by not having to work. So, counter-intuitively, women in wealthier families have less physical and economic autonomy.

India’s gender divergence (in sex ratios, employment, and autonomy) is clearly not a function of wealth. Rather, local gender norms mediate responses to economic growth.

Effect of colonialism

There are several ways in which colonialism might have compounded India’s gender divergence: via inheritance rights, progressive reforms, caste, land tenure or indirect/direct rule.

Inheritance rights

Some argue that colonialism compounded patriarchy by enabling Brahmin elites to codify Hindu law, which was then upheld by upper caste judges and had the net effect of curtailing female inheritance.

Before colonialism, disputes had been settled by local village or caste councils. Shastric prescriptions – concerning marriage, divorce and inheritance – were not necessarily practised by tribal communities or lower castes. Medieval temple inscriptions in Andhra Pradesh, Tamil Nadu and Karnataka suggest that women occasionally gifted land.

This implies female ownership. But, as Bina Agarwal notes, wealthy women’s pious acts could have just been a special category, exempt from patrilineal strictures. As she concludes, there is very little evidence to suggest Hindu women typically owned and controlled immovable property, before colonialism.

Brahmin interpretations of scripture varied geographically. Under the Bengal Presidency, they cemented Dayabhaga law (permitting widows’ inheritance). In Madras and Bombay, it was Mitakshara (proscribing widows’ inheritance). These regional differences long predate the Raj.

The colonial codification of Mitakshara could have worsened women’s inheritance rights in the South. But, that cannot explain why women now have more autonomy in the South.

Progressive reforms

Women’s bodies became a battleground during colonialism. British imperialists cast themselves as saviours, Indian liberals (like Gandhi, Jawaharlal Nehru, and Sarojini Naidu) sought social reform but under their control, while conservatives wanted to protect traditions from external attacks.

Female education was increasingly championed by educated, middle-class Indian nationalists. It symbolised respectability and refinement, without jeopardising women’s place in the home. Learned men published numerous critiques of polygamy, child marriage and purdah.

In the late nineteenth century, Indian liberals and women’s organisations campaigned for social reform. The Central Legislative Assembly passed emancipatory laws: prohibiting sati, child marriage, female infanticide; raising the age of consent and allowing widow remarriage.

These issues were all debated – irrespective of imperialism.

Feminist critique and mobilisation

The Women’s Indian Association was founded in Madras and the All India Women’s Conference was established in Maharashtra. Ten years in, 10,000+ members were organising for change. In Tamil Nadu, women joined the Dravidian movement and debated important social reforms.

Participants at the first Self-Respect Conference (held near Madras in 1929) demanded equal property rights for men and women. Their second conference pushed for female employment in the army and police. Women first won the right to be elected in Madras (1921). Bengali women agitated for the right to vote that year, but were defeated on the grounds this would extend suffrage to prostitutes. Women also joined the revolutionary struggle for sovereignty. Turnout was far higher in Bombay than Bengal. Again, feminist mobilisation was strongest in places that were already more gender-equal.

Role of Caste

The caste system influences gender relations in three important ways:

• Upper-caste purity and prestige has been preserved through female seclusion, prohibiting polluting sexual access – as highlighted by Uma Chakravarti. Upwardly mobile families gain status by following suit: curbing women’s independent mobility and pursuit of new economic opportunities. (Though there is a considerable jati-level variation).  
• Compliance is motivated by fears of social sanction. Men preserve their honour (izzat in Urdu) by policing female kin, for rumours of misconduct would soil the family name. Caste panchayats (assemblies of older men) are extremely powerful in rural areas, overseeing women’s sexuality and reproduction – as detailed by Prem Chowdhry. If a woman rejects her arranged marriage, the caste panchayat may severely fine her family or even outcaste them: prohibiting future marriages, cutting off their social networks and sources of mutual insurance. An entire lineage may be alienated and expelled from the village because of one daughter’s misdeeds. This heightens the costs of non-compliance and forestalls exposure to alternatives. Together with rural isolation, social policing limits exposure to more egalitarian alternatives.  
• Upper caste men’s political and economic dominance enables impunity for sexual violence against Dalit women – most recently in Uttar Pradesh.

Genetic data suggest that women’s sexuality and reproduction have been strictly policed by tightly-knit caste groups for millennia. Caste endogamy is truly ancient. The Vysya in Andhra Pradesh, for example, have been marrying within their caste, allowing no genetic mixing into their group, for over two thousand years. The Vysya have lived in close proximity to other castes, yet nonetheless maintained strict social isolation. This reflects a wider trend. So caste is not new.

That said, colonialism may have affected gender relations by enriching upper castes and compounding inequalities.

Land tenure

Colonial rule varied across India. Could this be the source of India’s gender divergence?

Banerjee and Iyer have attempted to categorise distinct judicial and administrative systems. In some parts of India, the British delegated authority to zamindar (landlords). Ever since the Mughals, the zamindar served as intermediaries: collecting revenue; controlling watchmen, police and courts. In other parts of India, the British sought to increase colonial coffers by taxing individuals directly or by vesting land rights in a group of villagers. Banerjee and Iyer find that in zamindari areas, the colonial state spent less on public goods.

But their crude schema is strongly contested. In practice, there seems to have been significant intra-regional variation. And with a more fine-grained, village-level analysis, these effects can disappear.

Colonialism does appear to have impaired governance in other ways though:

Weaker state capacity under colonialism is associated with fewer public goods today.

Direct colonial rule seems to have worsened outcomes. Areas formerly under native control have more schools, health centres, and roads in the postcolonial period.

The British also increased caste-inequalities in areas under their control: by granting property rights to landlords, reifying and ranking castes, as well as installing bureaucracies dominated by upper castes. Brahmins monopolised the highest offices under the Madras Presidency (just as they had served in the upper echelons of the Mughal regimes).

But these corrosive colonial governance regimes do not correlate with India’s gender divergence. Female literacy was highest across the South, notwithstanding differing degrees of imperialism.

In sum, there is very little evidence that colonialism contributed to India’s gender divergence:

• Even if elites entrenched patriarchal interpretations of scripture in Bombay and Madras, Southern women are still more autonomous than compatriots in Bengal
• Mobilisation and implementation of progressive reforms were strongest in areas that were already more gender-equal
• Castes have policed female sexuality for millennia, as revealed by genetics
• Direct colonial rule may have worsened caste-based inequalities, but this is not correlated with contemporary gender relations  
• Even if native rule improved public goods provisions, such as schools and clinics, access is mediated by pre-existing gender hierarchies (circumscribing women’s independent mobility). And where daughters are disposable, sonograms are just used to select male progeny.

Clearly, we need to go further back.

Matriliny cannot explain divergence

A few Indian communities are matrilineal: Khasis and Garos in the North-easterly hills, Nairs and Bunts on the southwesterly coast.

Men govern, but women remain relatively autonomous. They may move freely in their communities, enjoy pre-marital sexual freedoms, marry later, more easily divorce and often live in their natal village. With fewer strictures on their movements, Nair girls rushed to school and married later. Kerala led the way in female literacy.

But matrilineal communities are a minority and cannot explain India’s regional divergence. Tamil Nadu, Telangana and Andhra Pradesh women report even greater freedom of movement and labour force participation despite being patrilineal.

Cousin marriage and gender equality

Many South Indians idealise cross-cousin marriage. Southern women are more likely to be surrounded and supported by a familiar family. By contrast, Northern women marry outsiders, to become vulnerable strangers in their husband’s village – argued Dyson and Moore, in a famous paper with over two thousand citations.

Are you persuaded?

I am not.

Well, gender gaps in education are larger in communities where brides move out – as is common in the North. Perhaps parents invest less in daughters if they do not anticipate strong, enduring ties.

But neither East nor West India has much cousin marriage, yet their gender gaps in education are almost as small as the South’s. Cousin marriage might – conceivably – foster support for female inheritance, as assets remain within the male lineage. Indeed, Southern states were forerunners in permitting daughters’ inheritance rights and making women coparceners of joint family property. But these new laws actually exacerbated cousin marriage.

There is evidence that women from communities that allow intra-village marriage are more likely to move freely, travel alone, earn cash income and participate in self-help groups. But this is merely a correlation. As far as I am aware, no one has traced the causal process by which endogamy enhances autonomy. Something else in those communities may be advancing gender equality. And in practice, there is no correlation between a bride’s contact with her natal family and her proclivity to contribute to decisions, enjoy the freedom of movement, or access savings.

In 1901, fewer girls were missing from villages that extolled cross-cousin marriage. Perhaps parents did not resent dowry costs if they anticipated reciprocity and cost-sharing within the lineage. But there could be another explanation. Dowries were always common in Rajasthan, Bihar and Punjab but rare in Tamil Nadu, Karnataka and West Bengal. Lower-caste jatis in Tamil Nadu actually used to favour bride price. This reflects Southern women’s importance in wet-rice agriculture and mitigated the costs of daughters.

Indeed, there are many other features of Tamil Nadu, Andhra Pradesh and Karnataka that could have enhanced gender equality. In their absence of those conditions, cousin marriage does not seem to advance gender equality.

Consider the most patriarchal place in the world, the West Asia and North Africa. Cousin marriage is arranged by a third of West Asian and North African families. It ensures that female inheritance under Islam does not fragment patrilineal assets. Bound by cousin marriage, kinship groups share honour collectively. A woman’s impropriety shames the entire lineage. Women in the region are thus veiled, monitored and secluded. Cousin marriage reinforces kinship, limits her autonomy and is associated with low rates of female employment.

Cousin marriage is also practised by Muslims in North India and Pakistan, where women’s autonomy is strictly curtailed.

Southern women may have gained autonomy despite cousin marriage, not because of it.

A woman can be surrounded by kin but not necessarily more autonomous or better protected. Honour killings are a case in point – committed by brothers, uncles and fathers.

So personally I am not convinced that cousin marriage begets gender equality.

Conquests and purdah in North India

Purdah and honour cultures emerged in the harsh geographical terrain of North Arabia. Mountains, deserts and semi-arid steppes made settled agriculture impossible, except in the oases. Nomadic pastoralists moved in search of water, often fighting for new pastures, or allowing their animals to graze in a farmer’s field, and thieving their silver. Razzia (raids) were a constant threat. Arabia also hosted long-distance traders, connecting the Mediterranean, Africa, Mesopotamia and India.

Men were integral to pastoralism and trading: venturing on dangerous marches, handling camels and large oxen, often facing hand-to-hand combat.

In the absence of a strong state to police law and order, tight-knit kin vigilantly protected their assets. To preserve their honour, tribes often restricted women’s mobility. Veiling and seclusion symbolised wealth and status in pre-Islamic Byzantium and Sasanian Iran. Muslim rulers decreed wider adoption, veiling became integral to honour in large oasis settlements, where women likely encountered strangers.

Muslim regimes raided, pillaged and controlled India for over six hundred years. Mughal rule was concentrated in North India, on the upper Gangetic Plain. Rivers facilitated Akbar’s easterly conquests to Bengal’s fertile soils, while the Deccan constrained progress to the south.

Women were captured in raids. When Raja Dahir was killed in the 8th century, his wife and daughters were sent to Damascus as sex slaves. Affluent households, merchants and cultivators kept a few female slaves. Female slaves – as Ira Mukhoty details – were used as mules: farming, fetching water, smearing cow dung on the floor, disposing of human waste and for sex. If beautiful and/or talented, these women were sold as concubines for nobles. In the 10th century, the Rajputs of Rajasthan (who were subject to early attacks) started pracisting “jauhar” (women’s self-immolation to prevent military capture and preserve honour).

During Islamic rule, North Indian society became more gender-segregated. Since the ruling class practised purdah, it came to signify status. Upwardly mobile families followed suit, to symbolise respectability in an age of insecurity. New Hindu-Muslim converts were especially zealous in their performance of purdah. With Islamisation and the adoption of the plow, East Bengali women (once integral to wet-rice cultivation) slowly retreated to winnowing, soaking, parboiling and husking - within the confines of the family courtyard.

India’s caste-based society was already concerned by purity. If women were degraded by outsiders, male kin lost honour. Women concealed their bodies, lowered their gaze, averted their eyes, were chaperoned, and (if they could afford to) refrained from mixing with strangers. Segregation amplified gender inequalities. Female education dwindled. Men dominated the public sphere.

To preserve their purity and symbolise a ruler’s prestige, elite Rajput women were physically secluded. They are also absent from cultural representations. There are hundreds of portraits of Rajput noblemen – gifted to strengthen alliances, displayed to show the male lineage, and affirm men’s role in history. But there are no portraits of real, named Rajput women. Even when elite Rajput women commissioned portraits they did not do so of themselves. They upheld patriarchal norms. As art historian Molly Aitken reveals, elite Rajput women were made invisible.

Gender segregation persists through widely-shared expectations of social sanction. In the 1970s, fathers in rural Delhi feared that education could jeopardise their daughters’ marriage prospects. Other families might think she was no longer obedient. Girls themselves often envied peers who had the freedom to explore and learn about the wider world but could hardly go against their father’s will. On the Hindi belt, a bride expresses her resentment via song,

“O father you brought my brother up to be happy,
O father, you have brought your son up to give him your house,
And you have left a cage for me.”

In Benares in the 1980s, neighbours reported women’s improper conduct, telling relatives what they saw, scrambling her sisters’ marriage prospects. In rural Haryana, women who did not veil were often scolded, for it threatened family honour.

“Honour killings” occur when a woman’s impropriety disgraces her entire lineage. “Her action had soiled our honour,” explained Poonam’s father in North Delhi, after his brother had shot her in broad daylight.

Growing up, observing their families and communities, children learn that defiance is heavily punished. These patriarchal norms persist over generations, as parents teach their daughters to speak softly, show restraint, and respect elders. Even if Northern women complete secondary school, they are still less likely to choose their husbands.

Northern women’s autonomy is constrained through an arranged marriage, and the watchful eyes of joint families (which are more common in the North). Though young, professional women may wish to venture out, Northern cities are dangerous places.

In Delhi and Haryana, young women experience relentless sexual harassment – especially in overcrowded public transport and from unemployed male youth. Women fear for both their physical safety and their reputations – as observers see them going out and draw inferences about their impurity. Delhi (along with Bihar and Uttar Pradesh) consistently ranks as the most unsafe place for women. Fear of rape curbs female labour force participation across India. This effect is even greater for women who practice purdah/ are beaten for leaving without permission.

Female labour force participation is lowest on the Indo-Gangetic plain, where Muslim rule was concentrated. It is also much more gender-segregated. Most Delhi women working in manufacturing do so within their own home – with scant opportunities to expand their networks, organise, gain skills or autonomy. In Lucknow, working women are concentrated in subcontracted work and as unpaid labour in family enterprises. They work, but rarely interact with outsiders. They remain dependent on male intermediaries.

If women remain secluded, they are less likely to collectively critique and challenge their subordination. So women workers in Haryana do not always question gender wage gaps, for they presume men to be more competent. As a 19th century Haryana saying goes, “jeore se nara ghisna hai” (women as cattle bound, working and enduring all). In the Indo-Gangetic plain, most women eat after men have been served. This bias exacerbates sex ratios, via female malnutrition.

In sum, gender segregation became more widespread under Islamic rule. Men continue to dominate public life, while women are more rooted in their families, seldom gathering to resist structural inequalities.

But I must qualify the impact of the Islamic invasions.

First, even before the raids, purdah was observed by a few royal houses in the North. Second, other patriarchal practices like pre-pubescent marriage, proscriptions on widow remarriage and sati long predated the invasions. The Mughals actually criticised Sati, some even banned it. Akbar insisted on female consent, though others remained quiet as they feared revolt. Third, despite a shared culture of purdah, women’s labour force participation varies across the North. These differences may be rooted in traditional agriculture.

Female labour force participation

Before the modern era, almost everyone produced their own food, and the system for producing food was the most fundamental way in which gender ideologies became entrenched. Where women’s contribution to farming was relatively significant (shifting-cultivation and wet paddy fields), they have higher labour force participation today. Where men were integral to production (in wheat fields and plough-cultivation), women stayed at home. Over the centuries, gender divisions of labour became normalised.

In the forested hills of North East India women have always been integral to shifting cultivation. Women’s long-standing predominance in the public sphere has enhanced their physical and economic autonomy. Daughters are valued as providers, so sex ratios remain even.

Elsewhere in India, cultivation is less labour intensive, so women are not always needed in the fields.

Wheat has been grown for centuries on the fertile, alluvial Indo-Gangetic plain. Cultivation is not terribly labour-intensive, though cereals must still be processed, shelled and ground. This lowers demand for female labour in the field, and heightens its importance at home.

Rice-cultivation is much more labour intensive. It requires the construction of tanks and irrigation channels, planting, transplanting, and harvesting. Women are needed in the fields. Rice is the staple crop in the South.

Over the centuries, women’s work became normalised in rice-growing regions, and thus persists outside agriculture.

Urban female workforce participation is 11% higher in districts more conducive to rice rather than wheat-cultivation, under rain-fed and low-input conditions, finds Gautam Hazarika.

Soil texture varies across India. Southern districts have stickier, clayey soils. These are unsuitable for deep tillage. Farming is incredibly labour-intensive, with endless transplanting, fertilising, and weeding. These jobs are traditionally done by women.

Northern districts have more loamy soils, suitable for deep tillage. Men harness draft animals to prepare the land. This heightens the importance of male labour and lessens the need for (female) weeding. Eliana Carrenza finds lower female labour force participation and more uneven sex ratios in districts with more loamy soils.

Female labour market participation has fallen across India over the past three decades, but analysis by Lahoti and Swaminathan shows it has fallen the least in the South. This reflects women’s higher labour market commitment.

Type of work also varies across regions. “Contributing family worker” is the dominant type of work in all regions except Southern states, where women are more likely to work as casual labourers. Southern women are most likely to work for non-kin. This is consistent with women’s greater freedom of movement.

Over the centuries, Northern men’s roles as breadwinners became ingrained. Men went out to the fields while women remained at home. Thus even before the invasions, men may have been more important to agricultural production. Dowries are thus paid to the groom’s family. Daughters are an economic drain.

In other world regions where agriculture was traditionally male-dominated, women left family farms in search of new economic opportunities.

In the pre-industrial American North East, women and children were surplus to wheat production. Not needed at home, women responded to new opportunities in manufacturing. By 1832 over 40% of the industrial workforce in the American North East was young and female. Surplus female labour was similarly responsive to new opportunities in wheat-growing, medieval Europe. In slack periods, young women and men were a drain on resources. In England, only the firstborn son inherited. His brothers and sisters left to become hired labourers. Likewise in Latin America, women’s participation in farming is usually low, about 20%. They seldom inherit. Latin American women thus independently migrated to cities in search of jobs. In East Asia, women pursued factory employment to self-finance their dowries.

This occurred in the absence of social constraints: purdah, purity, and caste-based policing.

Semi-arid soils and sex ratios

In districts with historically low yields, girls are disproportionately likely to die. Hazarika, Jha and Sarangi have mapped ancestral yields per hectare, assuming it was rain-fed with low-input. Such districts are associated with worse sex ratios today - controlling for soil texture, religious and caste composition, monthly expenditure, and contemporary rainfall.

There is certainly a correlation between historically low yields and contemporary sex ratios. We can speculate several possible causal mechanisms.

Son preference is widespread across India. Sons are breadwinners, support elderly parents, perform ancestral rites and continue the lineage. As a popular saying in Haryana - recorded by Prem Chowdhry - goes, ‘Meehn aur bettya te koon dhappya sai’ (Who can be satisfied without rain and sons; both are necessary for cultivation). When resources are scarce, families prioritise sons. Strategic investment in sons may have been normalised through recurrent famines.

India is not unique in this regard. China is similarly patrilineal and patrilocal. Historically, when Chinese families struggled to survive (due to cholera, famine, or drought), they drowned girls at birth, or sold them as slaves/prostitutes/child-brides.

Perhaps these difficult decisions seldom arose in India’s South and North-easterly fertile soils? Given a more benign geography, letting girls die never became part of the culture.

An alternative hypothesis is that pastoralism was historically pervasive across North-west India, and this entrenched patriarchal norms. Pastoral societies tend to be gender-segregated. Men take the herd to pasture, while women stay at home, tending newborn animals and processing milk into ghee. Men may leave for a few days, searching for new pasture. If men cannot observe women’s whereabouts, they may worry about paternity, and try to control female sexuality.

Analysing societies across the world, Anke Becker finds that pre-industrial societies that were more dependent on pastoralism had stronger son preference and are more likely to believe in male superiority. These effects persist today. Women whose ancestors subsisted on pastoralism report less control over their sexuality and greater preference for sons, which is reflected in uneven sex ratios. Pastoral groups are also disproportionately patrilineal and patrilocal.

Rajasthan continues to be a major producer of livestock, wool, and dairy. Across the North-west, there are numerous pastoral communities (such as the Raika). Raika men head out, while women tend to veil their faces, eat after everyone else, and refrain from conversing with strangers - or at least in a low voice, from a distance. Jats (33 million strong, predominating in Punjab, Uttar Pradesh, Rajasthan, Delhi and Haryana) were historically pastoral.

In Indian districts with historically low yields, women are no less likely to work, but they are less likely to survive.

Arid Rajasthan exceeds the national average rate of rural female labour force participation. In Haryana, colonial officials observed that “women work as hard as the men if not harder”. Aside from ploughing, driving carts, or digging, there was no agricultural labour that a woman did not do. Women sow, weed, harvest, thresh, and maintain irrigation channels. But regardless of women’s contributions, men are prioritised.

Why has women’s importance in traditional agriculture not curbed son preference? Monica Das Gupta emphasises patrilineal, exogamous kinship: “perhaps the most important determinant of Punjabi parents’ attitudes toward girls is the fact that married women can do almost nothing for their natal kin”. She notes that many castes in North India will not accept food or water in their married daughters’ new home. Any hospitality must be generously paid for. Moreover, the lineage is entirely traced through the male line – as reflected in 19th century Rajput portraits. This institution of exogamous patrilineality is pervasive across North India. As Anke Becker shows, it is strongly associated with ancestral pastoralism.

Pastoralism may have also influenced India’s caste-system. Brahmins dominate business, public service, politics, the judiciary, and universities. Upper caste purity and prestige has been preserved through female seclusion, prohibiting polluting sexual access. These patriarchal norms may be rooted in ancient livelihoods. Brahmins share genetic data with ancient Iranians and steppe pastoralists. Brahmins also comprise a larger share of the population in North India and only 3% in Tamil Nadu.

Over the centuries, male superiority may have become entrenched. Generations of North Indian women have been breast-fed for shorter periods, given less nutritious foods, and tardily taken to clinics. Medieval Rajputs (predominating in North West India) highly extolled sati. “It is a small thing to kill a woman in an Indian village,” divulged a college-educated Jat, Punjab farmer in 1958. In arid Rajasthan and the surrounding deserts, women learn they are valued less.

Son preference persists, even as incomes rise in now-thriving Punjab and Haryana.

But ancestral crop yields only explain 12% of the total variation in sex ratios. They cannot explain why so many girls are missing on the fertile Indo-Gangetic flood plain. Additional factors contributing to son preference include patrilineal, exogamous kinship, men’s roles as providers (supporting elderly parents), and the high cost of dowries. Bride-givers in Rajasthan and Punjab traditionally paid dowries.

Dowries have escalated with economic growth and social stratification. Before 1930, only 38% of Indian households engaged in dowry-payments. By 1970, this had increased to 88%. The real value of dowry-payment tripled from 1930 and 1975 - calculate by Gaurav Chiplankar and Jeffrey Weaver. Southern families are increasingly paying dowries - to secure upward mobility. Husbands may even beat their new brides to coercively extract larger dowries, since she is unable to divorce. Many rural Tamil parents now perceive girls as economic burdens. This effect on sex ratios is compounded by pressure from low fertility and access to sex-selective technology – both of which are correlated with wealth. Thus, in some parts of Tamil Nadu, the sex ratio has actually worsened.

Even though dowry-payments have increased nationwide, dowry murders are clustered in North-central states: Haryana, Uttar Pradesh, Bihar. Control over dowries also varies geographically. In Tamil Nadu, brides often keep a portion for themselves, whereas in Uttar Pradesh, assets are usually handed over to in-laws – as documented by Srinivasan, Bedi, Jejeebhoy and Sathar. The North-South divide persists.

In sum

For centuries, Northern men have been fundamental to household survival, entitled to scarce resources and preserved their honour through female seclusion. This is a legacy of wheat-cultivation, deep-tillage, pastoralism, patrilineal patrilocal kinship, caste-based policing and invasions.

Northern parents increasingly support their daughters’ education, but this is primarily to improve their marriage prospects, not work outside the home. In Rajasthan, female labour force participation is relatively high in family farms, but very low in towns (where women would mix with non-kin). Ensuing gender segregation entrenches inequalities. It curbs exposure to women demonstrating their equal competence in socially valued domains and inhibits collective critique of patriarchal norms.

In Southern cities, women are visibly earning money, providing for their families. This is rooted in the historical absence of purdah and labour-intensive agriculture. Paid work is no panacea though. Given widespread condemnation of divorce and little independent property, wives may feel trapped in abusive relationships.

That said, by harnessing their social networks, Southern women have organised against discrimination: demanding dignity, safer cities and greater respect. In Mumbai 33 NGOs mobilised for “the Right to Pee”, advocating free, clean and safe toilets for women, asserting their right to public space.

Alice Evans is a Lecturer at King’s College London, a Faculty Associate at Harvard CID, with previous appointments at Cambridge and the LSE. This article first appeared on her website.

For thoughtful comments, critique and suggested readings, she is grateful to: Ajay Verghese, Arpit Gupta, Ananya Chakravarti, Bina Agarwal, Duman Bahrami-Rad, Gautam Hazarika, Keera Allendorf, Nathan Nunn, Vijayendra Rao and Pseudoerasmus.