If you are scratching your head thinking, what could possibly be a country with lower than usual sex ratios at birth (SRB) to complement the opposite of Azerbaijan’s skewed (high) SRB? I’m not going to be able to answer that to your satisfaction, because the answer is rather pedestrian. Lower sex ratios at birth show up here and there, but there is no real equivalent of son preference or sex selective abortion driving it. It’s mostly genetic predispositions and environmental harshness. So instead, I am going to the other end of the lifespan: The sex ratio at age 65+, using data from the UN World Population Prospects. I assumed the comparator would be a Baltic country. But, much to my surprise, it was Mozambique.
Source: UN World Population Prospects (2023)
And the surprises don’t stop there. When I mapped the data for sex ratios by age for the world (blue), Mozambique (green), and Azerbaijan (orange), it looks like yes Mozambique has the lowest ratio of females to males at age 65+ (about 50 men to 100 women), but the line ticks back up through the age progression. So this could mean that something specific happened to that cohort, or it’s an artifact of smaller sample sizes at older ages. Here’s to starting that investigation.
First: what explains the dips and dives in sex ratio across the life course?
I love this comprehensive yet pithy explainer from the US Census Bureau (from another partnership between USAID and other organizations that was beautiful). In short:
At Birth: While the natural baseline favors male births globally due to variations in fetal survival during gestation, wide regional differences exist. Cultural son preferences (which we talked about last week), resulting in prenatal sex-selection technologies or birth underreporting significantly skew birth ratios toward males in places like parts of Asia, whereas socioeconomic health disparities and genetics lean toward lower male-to-female ratios in Sub-Saharan Africa.
Infancy and Childhood: After birth, natural biological frailties typically cause higher male mortality rates (10-40% higher for infants). Historically, strong preferences for sons could reverse this trend through the neglect of female children, though the modern use of prenatal screening may replace this post-birth practice.
Working Ages: The excess of male deaths carries into adulthood driven by risky behavioral traits (like drinking and smoking) and dangerous occupational factors (like factory work or military service), balancing out the sex ratio around age 50. Conversely, geographic or regional sex ratios are highly sensitive to migration, shifting as men move to urban areas or international destinations for employment.
Older Population: Past the age of 50, sex ratios drop rapidly and become increasingly female-dominated. This stark imbalance is caused by escalating overall death rates combined with a higher male vulnerability to age-related health conditions, such as cardiovascular diseases and cancers, and female life expectancy that is on average, higher.
Other top skewed countries at older ages
I mentioned that I expected the Baltic states to top the charts for this indicator. My thinking was that two things could be at play here: World War II and alcohol-related mortality. And what do you know, Pew has the receipts: Yes, the gender imbalance originally spiked during World War II, when massive numbers of Soviet men died in battle or left to fight. By 1950, Russia had just 76.6 men per 100 women. Then, that number actually recovered for decades, climbing to 88.4 by 1995. But then, it reversed thanks to a second thing: alcoholism. Basically, younger men in this region die at an unusually high rate today, creating a massive 11-year gap in life expectancy, which demographers attribute to high rates of alcoholism (especially excess vodka use), suicide, and preventable diseases.
The UN data shows that Mozambique narrowly beats out Belarus, Russia, and Latvia here by a smidgen. But, while we saw this uptick after age 65 for Mozambique (previous figure), it is not the case for those countries. Which then makes me think that I could have used another age cut off, say 80+ and my country would have been different. We can have a whole different debate about using the 65 cut-off. I used it roughly as a proxy for retirement age, and because average life expectancy worldwide is roughly 74 (so wanted some years before that). Also, I will add shamelessly that at the time of writing this, I had done so much research already and bought all the ingredients for my Mozambique dish, so that’s what we will focus on this week 🙂
If you are curious though, here is what the sex ratio by age and sex looks like for Mozambique and the other top countries at age 65+. You can see that the Eastern European countries continue that decline, dipping to wildly skewed sex ratios at age 85/90 (!) OK, so what gives?
Mozambique, specifically: Birth cohorts matter
One of my favorite ways to visualize complicated annual data is to build birth cohorts where possible. In doing so for Mozambique, it’s clear that there is a generational story here. Using the UN data by year and age, we can create cohorts to show what trends in indicators, like sex ratio look like over time. Basically, folks that show up as 65+ in the year 2023 were born in or before 1958. Looking at their sex ratio at age 65 compared to the same ratio for folks born in other years is illuminating:
Essentially, sex ratios looked fairly normal up to a point. Then, two main things happened that shook the narrative.
The Mozambican civil war took place between 1977 and 1992 and was devastatingly brutal: there were roughly 1 million deaths in a country of 14 million, and about 5-6 million displaced. Basically, someone born in 1938 turned 39 the year the war began and 56 the year it ended. For that cohort - they were old enough to have mostly aged out of its worst exposure window by the time it hit the hardest. On the other hand, someone born in 1958, turned 19 when the war began and 34 the year it ended - squarely include the prime period for conscription, combat, and displacement.
The second gut punch for that cohort was HIV. The 1958 birth cohort turned 36 in 1994, which was the year that HIV prevalence in Maputo started doubling every two years. They were well within the age range that drove the epidemic, while the 1938 cohort mostly missed both catastrophes.
Are you shook? Because I’m shook.
A little note about the civil war/male mortality story: A detailed mortality study I came across actually shows that most recorded deaths (emphasis on recorded is important, if you have been paying attention to my rants about data and data quality) was highest among children under 5 (at least in Central Mozambique), and not men going to combat. The war’s role then was likely in the role it played in displacement, destruction of health and other infrastructure, and 15 years of social and economic chaos; which then set the table for a devastating impact of HIV once the war ended.
A closer look at HIV in Mozambique
There are readers here who can write this section better than I can, who worked tirelessly on extensive USAID HIV programming in the country, starting in 2003 with PEPFAR. In 2023, an estimated 2.2 million adults (15+ years) with HIV were living in Mozambique, which corresponds to about 12.5% of the total population.
The highest HIV prevalence is in a province called Gaza (20%), which if you recall the utter stupidity and chaotic destruction in 2025 when Gaza, Mozambique was confused with the other Gaza, with the Trump administration claiming that the US government is sending condoms to Hamas. Gaza’s HIV prevalence is 2x the national average (which is already one of the highest in the world), so it is a region with great need.
The reasons for the high prevalence are really interesting, so allow this minute digression: Southern Mozambique has long supplied migrant labor to South Africa’s mines, and remains one of the largest migrant groups there today. These migrant miners are separated from their spouses for months at a time, with documented higher rates of commercial sex work and alcohol use while away, thus carrying elevated HIV rest back home with them. We talked about birth spikes following sporting events (Week 24); institutional births in Gaza spike every September (9 months after miners return home for Christmas), and that same seasonal reunion is a documented spoke point for HIV transmission to partners back home.
The HIV prevalence is 2-3X higher among women aged 15-29 years compared to men, owing to age-disparate relationships with older male partners. So you would expect there to be more female mortality than male. Wrong! Once infected, men die at meaningfully higher rates. This is not because the virus operates differently in their bodies, but because of a well-documented gap in who actually gets diagnosed and treated. For example, a study from southern Mozambique quantified this:
Men were far less likely to get tested in the first place (48.7% to 62% of women);
Less likely to know their HIV status once infected (75.9% aware vs 88.9% for women);
4x more likely to arrive at their first clinical visit already in advanced stage AIDS; and
2x as likely to fall out of care after starting treatment.
2x more die in the 24 months after diagnosis.
Oof. And of course, this shows up in the sex ratio data at 65+.
What does this feminization of aging mean?
More women in their 60s and beyond have already outlived their husbands (often by decades), and are navigating old age largely alone.
There are structural inefficiencies in state social safety nets, such as the Basic Social Subsidy Programme (PSSB-Elderly), offer highly unequal regional coverage, leaving older women in the poorest provinces with minimal systemic support. So, state-sponsored safety nets are out; leaving familial ones as the only escape. We talked about rapid urbanization in Singapore and what it means for that country. Mozambique too is rapidly urbanizing, and it’s disproportionately young adults moving to cities, chasing work opportunities that do not exist in the countryside. For rural women in particular, this also means aging without a social and economic safety net of your children.
The practical costs are hardest on property. Under customary inheritance law, in much of patrilineal Mozambique, land, livestock and money pass through men; widows are often left without much of anything, and property grabbing by the deceased husband’s relatives is common. The feminization of aging in Mozambique is structurally bound to the feminization of poverty, creating a landscape of vulnerability for older rural women. As economic shocks, climate disasters, and inflation have pushed Mozambique’s general poverty rate toward an estimated 75% by 2025 (affecting roughly 1.35 million older adults), the burden falls disproportionately on women.
In some communities, remarriage after widowhood has been tied to pita-kufa, or “widow cleansing,” where a widow is expected to have unprotected sex with the brother of her deceased husband in order to remain under the family’s protection. There’s a lot to be said about this practice, but in this context, think about the feedback loop of HIV infection (likely causing her husband’s death, has likely infected her, and will now infect another).
And finally, there is an entire generation below. Mozambique has roughly 2 million orphans under 18, with Around 800,000 children have lost their parents to HIV and AIDS related illnesses. A lot of those children end up raised not by parents or other relatives, but their grandmothers. These women who have survived to older ages are now doing childcare for a second generation after burying a first.
Source: UNICEF
Looking ahead
Before 2025, the hopeful ending here would have been pretty straightforward: the generation now in their 40s and 50s will spend most of their adult years with HIV treatment (ART) available, affordable, and accessible. This should have eventually produced a 65+ ratio that climbs back towards the high-80s where the pre-1938 cohorts once sat.
UNAIDS set the 95-95-95 targets in 2021 to end the epidemic by 2030 by which:
The progress made needs funding to remain sustainable through these goals (and beyond). PEPFAR (from the US Government) provided around two-thirds of the funding for the country’s HIV program for more than a decade.
The US government announced a pause in all foreign assistance through USAID on 21 January 2025. USAID grant holders were told to stop work on 24 January and thousands of HIV testing counsellors and clinic staff were laid off in programmes supported by USAID funding. In some countries, vital medications were stranded in warehouses because USAID funded the supply chain that brought those drugs to pharmacies. Although a limited waiver was granted to allow funding of PEPFAR treatment and care services, studies from Mozambique and South Africa’s largest city show the swift impact of the USAID funding cut on testing, HIV diagnosis and new treatment initiations in the first quarter of 2025. - AIDS map news (2025)
A team of researchers estimated that 83,000 new HIV infections would occur in Mozambique by 2030 as a result of the loss of funding, a 15% increase compared to the previous epidemic trend. An additional 14,000 people would die of HIV-related causes, a 10% increase over the previous trend.
What happens next is something that we will have to wait and watch. Whether the funding gap will be filled by other donors, the Mozambican government, or something else entirely. Instead of stabilizing, perhaps the next cohort may end up showing this skewed sex ratio as well, but just hit by this new third mechanism than the generation before.
The Food!
I know you are expecting some type of peri-peri situation. I was, too. But then I saw this recipe for a shrimp curry and my genes couldn’t help themselves. We made Caril De Camaro (STUNNING!), coconut rice, and this avocado+peach salad that was a banger. We listened to music by Neyma, and had a long conversation about colonization.
Fascinating read and thanks for the music tips! Would love to be a fly on the wall to listen in to your conversation, esp what the kids have to say
Great use of birth cohorts to explain the skewed sex ratios. Going off now to find some coconut rice.