Counting what surveys miss

#In rural Peru, a ballot box reveals violence that standard surveys miss

An NBER working paper released on October 12th randomly assigned 5,562 women in rural and peri-urban Peru, surveyed in 2018, to answer questions on violence by their partners aloud to an interviewer, on a private ballot or through a list experiment. The ballot raised reported rates of any violence by 17%, with significant differences on six of seven questions.

Why it matters: Underreporting is not random: it is worst among women with little schooling and for episodes linked to cash on hand. With ballot data, reported sexual violence roughly doubles, to around 20%, after Juntos transfers arrive; in face-to-face answers it stays near 10% throughout the payment cycle. Evaluations built on standard questions may therefore misread what drives violence; this week's Perspectives takes up the question.

Source: J. M. Agüero, U. T. Aldana, E. M. Field, V. Frisancho, I. Rodriguez Hurtado and J. Romero, NBER Working Paper 35891, October 12th 2026 · NBER w35891

#In Mexico, dry spells raise recorded violence against women

A UNU-WIDER working paper by Sandra Aguilar-Gomez and Andrea Salazar-Diaz, registered on October 12th, links unexpected days without rain to three records of intimate partner violence: injuries in the public health system, police reports and 911 calls. All three rise after dry spells, most in regions that depend on agriculture and during the growing season.

Why it matters: The proposed channel is work. Women in farming lose hours and find no other income, while men partly move into off-farm jobs; less income weakens women's bargaining position, and more time at home raises their exposure to an abusive partner. Because the evidence comes from administrative records rather than surveys, it is a useful check on survey-based findings.

Editor's note: summary based on the published abstract; we have not read the full text.

Source: S. Aguilar-Gomez and A. Salazar-Diaz, WIDER Working Paper 113/2026, October 2026 · UNU-WIDER · doi:10.35188/UNU-WIDER/2026/788-6

#In India's joint households, the daughter-in-law does the housework

Using India's Periodic Labour Force Survey and Time Use Survey, Tanu Gupta and Digvijay Negi compare daughters-in-law with unmarried daughters of similar age and education living in the same household. Daughters-in-law spend about 245 more minutes a day on domestic work and care, and are less likely to hold paid jobs.

Why it matters: India's low female labor-force participation is often attributed to norms in general; this paper locates the constraint in a specific position in the household hierarchy. The design is descriptive, as the authors stress, so the gap is not a causal effect. It still argues for measuring time and agency person by person rather than household by household.

Editor's note: summary based on the published abstract; we have not read the full text.

Source: T. Gupta and D. S. Negi, Development Policy Review, October 12th 2026 · doi:10.1111/dpr.70107

#Brazil's university quotas: large gains for beneficiaries, smaller losses spread wider

An NBER working paper by Nano Barahona, Cauê Dobbin and Sebastián Otero, released on October 12th, studies Brazil's rule requiring federal institutions to reserve half their seats for racially and socioeconomically disadvantaged applicants. At admissions cutoffs, students who gain a seat attend and graduate from better programs; their graduation gains are about twice the losses of the students they displace.

Why it matters: The catch is the chain of reallocations: the policy displaces about 30% more students than it benefits, so losses offset about two-thirds of the gains, leaving a modest net gain. The estimates combine comparisons at admissions cutoffs with a structural model of degree choices for the overall effects. It is a rare full accounting of who gains and who pays when access to higher education is redistributed.

Editor's note: summary based on the published abstract; we have not read the full text.

Source: N. Barahona, C. Dobbin and S. Otero, NBER Working Paper 35893, October 12th 2026 · NBER w35893

#Mobile internet raises employment in poorer places, mostly in vulnerable work

An ILO working paper published on October 8th combines labor-force survey microdata and coverage maps for 52 countries, covering about 28% of the global labor force, from 2004 to 2023, and uses regional lightning density, which affects network rollout, as an instrument for 3G expansion. In lower-income, low-coverage settings, connectivity raises participation and employment, with stronger gains for women; in richer, more connected ones, it lowers participation.

Why it matters: In poorer places the gains come largely through own-account and family work and agriculture, not wage jobs. And 3G does not systematically narrow gender gaps: young women in lower-income countries become relatively more likely to be neither employed nor in education or training. The sample is mostly upper-middle- and high-income countries, so lessons for the poorest need care.

Source: P. Carrillo, M. Golin, D. Bescond, A. Mahajan and R. Viné, ILO Working Paper 180, October 8th 2026 · ILO · doi:10.54394/00034480

#In Sub-Saharan Africa, drought lengthens women's walk for water

Xueyang Wang and Wenling Liu combine Demographic and Health Surveys with geocoded drought exposure. Drought is associated with more time spent collecting water by women, especially in poor households, among women with no formal education and where there is no piped water, and with more trips longer than 60 minutes.

Why it matters: Higher community water burdens are also associated with girls being out of school and further behind for their age. These are associations, not causal estimates. They point to a cost of drought that income and production statistics miss, paid in women's unpaid time and girls' schooling.

Editor's note: summary based on the published abstract; we have not read the full text.

Source: X. Wang and W. Liu, World Development, October 9th 2026 · doi:10.1016/j.worlddev.2026.107592

#Mozambique's sugar collapse cut schooling most for children of less educated parents

A UNU-WIDER working paper by Giulia Barletta and Sam Jones uses the Vulnerable Lives Survey 2024, Mozambique's first retrospective intergenerational survey, and the 1997 census to study the collapse of the country's sugar-estate sector. Greater exposure in childhood lowers educational attainment and raises the probability that children fail to surpass their parents' schooling.

Why it matters: The largest losses fall on children from less educated families, so the shock hardened inherited disadvantage rather than leveling it. That is relevant wherever a dominant employer or sector fails, from mining towns to commodity regions. The abstract reports no magnitudes; the size of the effect needs the full paper.

Editor's note: summary based on the published abstract; we have not read the full text.

Source: G. Barletta and S. Jones, WIDER Working Paper 114/2026, October 2026 · UNU-WIDER · doi:10.35188/UNU-WIDER/2026/789-3

#In Loreto, Peru's ombudsman counts over 218 mothers younger than 14

On October 12th Peru's Defensoría del Pueblo reported more than 218 cases this year of girls under 14 who have become mothers in Loreto, which ranks first nationally according to the health ministry's online birth-certificate registry. It also cited more than 3,466 pregnancies among girls under 18 in the region, without stating the period.

Why it matters: Pregnancy at that age signals sexual violence and closes off schooling and choices before they begin. The ombudsman says births in remote Amazon communities often go unregistered, so the count is a floor, and that the region lacks Gesell rooms, used to take children's testimony without revictimizing them. The figures are administrative counts, not survey estimates.

Source: Defensoría del Pueblo, Nota de Prensa 630/OGCO/DP/2026, October 12th 2026 · Defensoría del Pueblo · NP 630 (PDF)

#The IMF expects Bolivia's economy to shrink by 3% this year

Projections for the IMF's October 2026 World Economic Outlook, released for the Annual Meetings, have Bolivia's real GDP falling by 3% in 2026 and 1.5% in 2027, after 1.6% in 2025. It is the only South American economy projected to contract in either year; Peru is expected to grow by 3% in 2026.

Why it matters: The forecast is slightly worse than the World Bank's 2.8% contraction of the previous week. The Fund projects consumer-price inflation of 14.2% by the end of 2026, after 20.4% at the end of 2025, and unemployment rising from 3.3% to 4.5%. Higher joblessness on top of still-high inflation is the combination that tests social protection during the adjustment.

Source: IMF, World Economic Outlook database, October 2026 · IMF DataMapper · Visión 360

#Bolivia sets a monthly allowance for children orphaned by femicide

Supreme Decree 5745, published in the Gaceta Oficial on October 11th, regulates Law 1680 of November 2025. Children orphaned by femicide and other killings will receive a monthly payment equal to 20% of the national minimum wage, paid through the public bank without deductions or fees, up to age 25 for those in education; those under five also receive a monthly food package.

Why it matters: The program is small but precise: a defined amount, a named payer and a ban on deductions. The presidency estimated, as reported by EFE, that some 961 children and adolescents had been orphaned by femicide by December 2025. Eligibility requires registration with departmental social-policy offices and documents from a criminal case, a hurdle worth monitoring for the relatives who care for these children.

Source: Decreto Supremo 5745, Gaceta Oficial del Estado Plurinacional de Bolivia, October 11th 2026; EFE, as reported by Correo del Sur · Correo del Sur (EFE)

#Peru drafts school-discipline rules that allow permanent expulsion

On October 12th Peru's government presented draft guidelines for managing discipline in basic education, pre-published by Ministerial Resolution 613-2026-MINEDU for 15 days of public comment. They would apply to public and private schools, and pupils who commit very serious offenses would be permanently separated.

Why it matters: School is the public service that most shapes later opportunity, so exclusion is a high-stakes sanction. The official newspaper's report does not say whether expelled pupils keep a right to a place elsewhere in the system. That, and who decides what counts as very serious, are the details to read in the draft during the comment period.

Editor's note: summary based on the report in El Peruano; we have not read the draft text.

Source: Ministry of Education (MINEDU), as reported by El Peruano, October 13th 2026 · El Peruano

#Over 14 years, 37% of Peru's state-funded scholars abroad did not return

Pronabec, Peru's national scholarship agency, said on October 13th that 63% of the scholars who studied abroad in the program's 14 years returned to Peru, and that just over half of those fulfilled their commitment to serve in the country. It plans 100 scholarships abroad next year, against the usual 150, and to expand one technical-training scholarship from 300 to 3,000 places.

Why it matters: Public money for study abroad buys opportunity for a few; whether it comes back as skills at home depends on enforcement and on the jobs that await returnees. Moving places toward technical training changes who benefits. The figures are the agency's own, reported by the state news agency, not an evaluation.

Source: Pronabec, as reported by Andina, October 13th 2026 · Andina