Thimphu Housing: Save Twice What You Earn

A three-bedroom apartment recently listed in Changzamtog is priced at Nu 7.5 million. To buy it within ten years, a household would need to put aside Nu 62,500 every month.

The median urban household in Bhutan earns Nu 28,000 a month.

So the arithmetic is not that the flat is expensive. The arithmetic is that a median Thimphu household would need to save more than twice its entire income, every month, for a decade — and that is before interest, before the deposit, before living.

This is not a difficult purchase. It is an impossible one, and the impossibility is arithmetic rather than rhetorical.

What the world’s yardstick says

There is a standard measure for this, and Bhutan uses it.

For rent, the international benchmark is 30% of household income. It comes from American public-housing law — set at 25% in 1969, raised to 30% in 1981 — and is now near-universal: above 30% a household is “cost burdened,” above 50% “severely cost burdened.” Australia applies the same 30% line. Bhutan’s National Housing Policy 2020 adopts exactly that threshold.

Thimphu tenants spend about 42%.

That is not an outlier reading. A 2022 rent study put median Thimphu rent at Nu 8,000, or 37% of median household income, and 48% in Hejo-Samtenling. The trend has been measured for years and it has been moving in one direction.

So by its own government’s standard, Thimphu is running 40% over the line.

The allowance that stopped in 2015

Here is where the state’s position becomes hard to defend.

Civil servants receive a House Rent Allowance, introduced in 2015 and set against the rental market as it stood then — 20% of the minimum revised pay scale, with staff at supervisory and support level 3 and below receiving a flat Nu 3,500 a month.

It has not been changed since. Not once in eleven years. It has never been systematically reviewed against rents or against inflation.

One- and two-bedroom flats in Thimphu now routinely rent above Nu 12,000. The state’s contribution toward housing its own employees therefore covers something like a quarter of what housing actually costs. A senior civil servant, quoted by Kuensel, put the gap precisely: had the allowance tracked the rent-escalation framework in Bhutan’s own Tenancy Act, it would be more than 60% higher today.

The government acknowledges that rents have risen. It has committed to no review, no timeline and no figure.

The deposit is the door, and it is shut

The financing picture explains why so few households cross from renting to owning.

Only about 9% of households hold a housing loan. Rates run 8.13% for non-commercial housing and 8.93% for commercial — among the highest in the region — while housing microfinance charges 10 to 18% and remains small.

Banks will lend up to 80% of a property’s value over as long as thirty years. Which sounds accommodating until you notice what it means: the borrower must produce the other 20% in cash.

On that Changzamtog flat, the deposit alone is Nu 1.5 million — fifty-four months of a median household’s entire income, saved in full, before a single ngultrum is borrowed.

The mortgage is not the barrier. The deposit is, and no interest-rate policy touches it.

Against the cities they are leaving for

The most useful comparison is not Bhutan against itself over time. It is Thimphu against the places Bhutanese are actually moving to.

Measured as the price of a dwelling divided by annual household income — the standard international ratio — Thimphu comes out between roughly 13 and 22, depending on an assumption I will be explicit about in a moment.

Perth is 8.23. More Bhutanese live in Perth than anywhere else outside Bhutan. On every version of the Thimphu calculation, housing in Perth is more affordable relative to local incomes than housing in Thimphu — and the same holds for Melbourne, Auckland and Sydney.

That is worth sitting with. The pull of Australia is usually described as a wage story. It is also a housing story: the wage buys a home there and does not here. A nurse who leaves Thimphu for Perth is not only earning more, she is moving into a market where the ratio between what she earns and what a flat costs is roughly a third of what she left behind.

What I do not know, and it matters

That 13–22 range is wide because the calculation has a defect I would rather state than bury.

The prices are 2026. The income figure is from 2022 — the Bhutan Living Standards Survey, which is the most recent published measure of median urban household income. In between sits the 2023 civil-service pay revision, the largest in Bhutanese history for some grades. If urban incomes have risen materially since 2022, the true ratio is lower than the headline arithmetic suggests: at unchanged incomes it is 19–22, at incomes up a third it is around 15, at incomes up half it is closer to 13.

At the top of that range Thimphu is less affordable than any city in Europe or North America. At the bottom it sits alongside Sydney and Zurich. The honest position is the range, not the flattering end of it.

A more recent income figure would settle it, and here is the difficulty: there isn’t one. The Living Standards Survey runs roughly every five years — 2017, then 2022 — so the next reading is not due until about 2027. Bhutan will spend the intervening years unable to say, from published data, whether its capital city’s housing became more or less affordable during the largest pay revision in its history. That is not a gap in this article’s research. It is a gap in the national statistics.

Two things survive the uncertainty in any case. Thimphu is dearer than Perth, on every assumption. And Thimphu is considerably more affordable than Kathmandu at 32.6 — the Bhutanese housing problem is severe when measured against rich countries and unremarkable when measured against its neighbours. Both are true, and a piece that reports only the first is not reporting.

The same measurement, at the other end of the scale

This site has an index for the smallest purchase a salary makes. The Momo Index prices the ngultrum in plates of momo: a graduate’s first month’s pay bought around 425 plates two decades ago and buys closer to 280 now — an erosion of about a third, because street food has run at roughly twice general inflation.

Housing is the same measurement at the opposite end, and it behaves far worse. A plate of momo went from about Nu 20 to about Nu 120 across some twenty-three years, near 8% a year. Land in Babesa went from Nu 300,000 a decimal in 2015 to between Nu 900,000 and Nu 1.5 million by 2021 — 20 to 31% a year, over a much shorter window. Land has been running at two to four times the momo rate, and momo already runs at twice general inflation.

The bigger the purchase, the faster it escapes the wage. That is why the two measures reach different verdicts about the same country. A third fewer plates of momo is real and survivable. A home priced at 13 to 22 years of household income is not a hardship; it is the removal of a plan.

Which is the honest answer to why people leave. Nobody emigrates over lunch. But the momo stall is where the arithmetic gets noticed daily, and the apartment listing is where the conclusion gets drawn.

One caution about our own earlier number, while we are here. The Momo Index offers a more conservative reading beside its headline — erosion nearer a fifth measured against the whole cost of living rather than a third measured in momo. That conservative figure leans on the consumer price basket, in which the non-food housing weight of 21.7% is of unclear construction: whether it tracks actual rental transactions or imputed owner-occupier costs is not specified. If housing is the fastest-moving component and also the most uncertainly measured, the careful one-fifth probably understates the erosion, and the vivid one-third may be nearer the truth than we allowed.

The supply that was not built

Against all this, delivery.

Sixty-four affordable housing units were completed in the last five years. That figure was given to the National Council in June 2026, by the minister responsible.

At the same hearing came the reason, and it is not obscure. Nu 1.5 billion allocated from the Economic Stimulus Package for affordable housing was redirected to the fuel subsidy when global prices escalated. The money existed. It went to hold the pump price down, and the housing was not built.

Meanwhile the market that does exist is only partly visible. The Department of Revenue and Customs recovered Nu 66.38 million over five years from undeclared and under-declared rental income — Nu 21.89 million in tax and Nu 44.49 million in fines, penalties running at twice the tax itself. Cash transactions, informal arrangements, deliberate non-compliance. Whatever the official rent statistics measure, it is not quite the market.

Bhutan has been here before

One last thing, because it complicates the obvious remedy.

In March 1988 a consolidated-salary rule fixed rent at 30% of salary — the same threshold the National Housing Policy uses today, applied as binding policy rather than as a benchmark.

Between 1988 and 1991, eight to ten buildings went up in Thimphu. In the comparable earlier period, sixty-seven.

That is not an argument against tenant protection. It is a warning that the easiest lever — cap the rent — has been pulled in this city before, and what it produced was not affordability but an absence of buildings. The problem is on the supply side, and sixty-four units in five years is the measure of how little has been done about it.

What the arithmetic actually says

The aspiration has already downshifted once. A generation ago the expectation was a house; today it is a flat, and people describe even that as luck rather than plan. Families stay in rented apartments far longer than intended. Adult children stay in the parental home. Household formation is being delayed by arithmetic.

And the downshift did not help. The compromise costs 13 to 22 years of total household income, requires a deposit of four and a half years of income before borrowing, and is supported by a state allowance last adjusted when the flat cost a fraction of today’s price.

The question this poses is not whether Bhutanese housing is expensive by world standards. Measured against Kathmandu or Bangkok, it is not. The question is why a country where a two-bedroom flat costs USD 80,000 has made that flat unreachable for the household earning the median wage — and why, over eleven years in which rents roughly doubled, the single number the state controls outright has not moved by a ngultrum.