Research

Since large-scale mining began in 2004, Mongolia's economy has grown about 7.2% a year. Mining now supplies roughly a quarter of GDP, 90% of exports and more than 30% of government revenue. Two of the world's largest deposits, Oyu Tolgoi (copper and gold) and Tavan Tolgoi (coal), sit in Umnugovi province in the South Gobi, and the companies developing them have promised money for local schools and infrastructure: Oyu Tolgoi's 2015 cooperation agreement with four partner districts committed five million dollars a year to a community fund, and a new school and kindergarten complex opened in Khanbogd in 2018. At the same time, UNICEF found schools in South Gobi mining towns running at 50 to 250% of capacity, with the Khanbogd school ranked last among 21 in the province, and the ILO put about 56,000 Mongolian children, a tenth of 5 to 17 year olds, in child labor, with mining among the hazardous sectors that draw in teenage boys. A 2021 to 2022 national survey counted 138,525 children in child labor, 16 percent of the age group, most of it herding and fetching water and fuel in rural households.

How has large-scale and informal mining affected education and the economy across Mongolia's 21 provinces, and what happened to the existing school systems in mining towns once companies arrived and families migrated to work?

Education and economic data from Mongolia's National Statistical Office were combined for all 21 provinces from 2000 to 2022 (enrollment by level and gender, graduates, teachers, schools, kindergartens, per capita GDP and industrial sales), with district-level data from 2015. A 2022 study identifying the four provinces where mining measurably raises household income (Tuv, Selenge, Orkhon and Umnugovi) defined the mining group; the other seventeen served as the comparison. Outcomes were compared before and after mining activity began, then the four mining provinces were split by type: large-scale corporate mining (Umnugovi, Orkhon) against informal small-scale gold mining (Tuv, Selenge). A third test looked inside Umnugovi at the districts covered by the Oyu Tolgoi agreement.

Taken together, mining provinces showed no significant change in enrollment. The result appears only when the provinces are split by the kind of mining. Where mining is large-scale and formal, enrollment rose in primary and secondary school for both boys and girls, along with per capita GDP and industrial sales, while the number of schools and kindergartens fell: more children in fewer classrooms. Where mining is informal, the opposite: enrollment fell at every level, fewer students finished grades 9 and 12, and per capita GDP fell. In the Oyu Tolgoi partner districts, the dependency ratio and truck traffic rose within a year of the agreement, but school entrants, students and schools did not measurably change.

The effect of mining on a province's schools depends on what kind of mine it is. A large corporate mine brings revenue and, on the numbers, more children enrolled, but it also brings unregistered in-migration that overloads the schools that exist. Informal mining pulls parents and teenage boys into unregulated work and pulls enrollment down. Limits: district-level data only begins in 2015, mining-town populations are undercounted, and some provinces host both kinds of mining at once.

Provincial figures for 2023 to 2025, retrieved from the National Statistical Office for the extended report, show the first half of the pattern persisting: the large-scale provinces' enrollment lead over the comparison provinces widened from about four points in 2022 to about six in 2025, and their student numbers have grown by about a third since 2019 against 15 percent in the comparison provinces. The second half has weakened: the informal provinces' enrollment deficit, five to eight points in 2015 to 2018, had closed by 2025, driven by Tuv. Whether that reflects the formalization of artisanal mining or the re-basing of population estimates after the 2020 census, the data cannot say.

Mongolia moved from a herding economy to a mining economy in a single generation. After the collapse of the socialist system, herds grew from 26 million animals in 1990 to a peak of 71 million in 2022, overgrazing and brutal winters followed, and the capital's population more than doubled, from about 780,000 in 2000 to 1.8 million in 2025, roughly half the country. Mining filled the gap: GDP growth averaged over 10% a year between 2011 and 2015, foreign investment reached 2.4 billion dollars in 2019, and Rio Tinto's Oyu Tolgoi mine, which produced 345,000 tonnes of copper in 2025, is expected to reach about 500,000 tonnes a year from 2028.

The same industry runs on conditions the country has not been able to regulate. A 2008 ILO study of 232 children working at informal mines found 56% of them underground and 53% injured on the job; the fine for employing a child was then under thirty dollars, and the 2021 labor law sets a minimum age of 15 but, according to the US Department of Labor, no penalty for breaking it. Herders near mines report contaminated water and lost pasture. Tsetsegee Munkhbayar, the herder who founded the Onggi River Movement and won the Goldman Environmental Prize in 2007, was sentenced to 21 years for terrorism in 2014 after an armed protest outside parliament; the sentence was cut to seven years on appeal and he was released under an amnesty in 2015. Company disclosures show the investment (20.6 billion dollars spent in Mongolia since 2010, 6.1 billion dollars in taxes and fees, about 17,000 employees) and not the rest. And by the World Bank's 2020 count, only one cent of every mining dollar earned over the previous twenty years had been saved for future generations.

Eliminating mining is not a realistic option; the question is what makes it survivable. Three approaches are already in motion. Public knowledge: the World Resources Institute's toolkit for mining-affected communities has helped residents get licenses revoked, agreements renegotiated, and local projects funded, which shows that communities can advocate for themselves once they have the information. Structural reform: the World Bank's four recommendations, countercyclical fiscal policy to smooth the boom and bust, investor rights in a competitive market so industries other than mining can grow, better use of a young and educated workforce, especially women, and regulation to reduce corruption, a shift they call 'from mines to minds.' Saving the revenue: a sovereign wealth fund law passed in 2024 is the first structural attempt to keep mining income for the future rather than spend it in the year it arrives; its first payout, in 2025, credited about 36,000 tugrik to every citizen's account, usable only for health, education and housing. And in June 2026 the government and Rio Tinto agreed to adjust the interest on Oyu Tolgoi's shareholder loans, a change the Prime Minister valued at 6.2 billion dollars, and to bring forward Mongolia's dividends, which had been projected for 2041.

Each approach depends on the others. Community pressure needs transparent data to work with, transparency needs a government willing to enforce it, and enforcement needs a budget that does not itself depend on the companies being regulated. That circularity, not any single policy, is the problem.

Ulaanbaatar is the coldest capital in the world, and about half of Mongolia's population lives there. Between 800,000 and a million people, more than half the city and over a fifth of the country, live in unplanned ger districts on its edge with no heating infrastructure. Through six-month winters some 200,000 households burn roughly 600,000 tons of solid fuel, and studies attribute between 56 and 80% of the city's winter fine-particle pollution to their stoves. UNICEF put the average PM2.5 concentration in 2016 at 256 micrograms per cubic meter, 25 times the WHO guideline of the time, which has since been halved. UNICEF called it a child health crisis: respiratory infections up 2.7 times in a decade, pneumonia the second most frequent cause of death for children under five, and children in central Ulaanbaatar showing 40% lower lung function than rural children. The annual health cost is estimated at 177 to 727 million dollars.

A decade of policy preceded the current approach: the 2010 law on air pollution fees and the 2012 Law on Air, the World Bank-backed Clean Air Project subsidizing cleaner stoves and electric heating, the 2016 Law on Hygiene banning waste burning, distributed face masks, and about 150 billion tugrik of public spending plus 60 million dollars of foreign aid between 2008 and 2018. In May 2019 the government banned raw coal for household use and replaced it with refined briquettes subsidized to the same price.

Partly. The briquettes burn longer and emit less, and winter PM2.5 fell by about 40% in the first season. Annual averages fell from 62 micrograms per cubic meter in 2019 to 26 in 2024, still five times the WHO guideline, and on a February morning in 2026 the city again topped the world's pollution rankings. The switch also carried a cost the policy did not price: carbon monoxide poisonings in the ban districts rose about six-fold, with 91 deaths in the three winters after the ban against 33 in the two winters before. But a subsidy priced to match raw coal still ignores the true cost of burning it, so the market failure the policy set out to correct remains. And the burden fell on the poorest households: when COVID cut incomes, families who could not afford briquettes burned trash and other flammables instead. The ban was more efficient than what came before it and less equitable than it needed to be. A policy judged only by the city's air quality reads as a success; judged by who paid for it, less so.

Mongolia's average temperature has risen 2.2 degrees since 1940, well above the global average, and the winter disasters known as dzuds, once roughly a once-a-decade event, have struck in one year out of five since 2000. The World Bank expects one every three years by mid-century. The 2023 to 2024 winter killed 7.9 million animals, and the national herd fell from a record 71 million in 2022 to 58 million in 2024. Two-thirds of the rangeland is degraded. Every severe winter pushes households toward the mining towns and the capital, and their children arrive in schools that were not built for them. The migration that the first study traced to the winters of 1999 to 2002 is now a recurring shock rather than a single event.

Coal was 12% of GDP and 58% of exports in 2023, almost all of it sold to China, whose steel output has fallen for two years running and whose demand for coking coal the International Energy Agency expects to decline through 2030. Mongolia's 2025 experience, record export volumes at a third less revenue, is what that looks like from the supplier's side. Copper is the hedge: Oyu Tolgoi is expected to reach 500,000 tonnes a year from 2028, the IEA projects a global copper shortfall by 2035, and the June 2026 agreement brings Mongolia's dividends forward. But copper mines draw migrants and water the way coal mines do, and only 12 of the 75 herder wells that the Oyu Tolgoi expert panel recommended in 2017 had been built by 2024.

Public consumption and investment both rose by more than 45% in 2024, and the IMF projects public debt rising from 45% of GDP in 2025 to about 64% by 2031 under current policies, above the country's own 60% ceiling. The sovereign wealth fund created in 2024 exists mostly on paper so far: the Future Heritage Fund is frozen until 2030 and the Savings Fund has been lent to banks. The pattern the World Bank described in 2020, 99 cents consumed and one cent saved, has not yet been broken, and the royalty collapse of early 2025 showed how quickly a bust arrives.

Enrollment near 100% is a solved problem in Mongolia; learning is not. In PISA 2025, Mongolian fifteen-year-olds scored 419 in mathematics, 374 in reading and 424 in science against OECD averages of 463, 461 and 482, and 43% were below the baseline in science. Primary numeracy proficiency fell from 54% to 38% between 2019 and 2021. Life expectancy is 68 years for men and 77 for women, and air pollution is attributed more than 1,000 deaths a year in the capital. Where mining is formal, children are in school, but the household evidence suggests they are not completing or learning more; where it is informal, for at least part of the period, enrollment ran five to eight points below the comparison provinces. Neither produces the workforce that a shift from mines to minds assumes.