The Frozen Pension Arithmetic

Two people retire in the same year. Same employer history, same National Insurance record, same claim date, the same full new State Pension down to the penny. One settles in Chiang Mai. One settles in Cebu. A decade later, one of them is receiving £241.30 a week and the other is receiving £155.65 a week, and the only variable that produced that gap is which country’s name is on the address the Department for Work and Pensions has on file.

That is the whole subject. Not a risk, not a projection, not a worst case. A rule, already in force, already costing a specific number of pounds to a specific person, every week, while the cheerful relocation content for both countries quotes the same starting pension and stops there.

The rule nobody is shown

The UK State Pension is payable anywhere. It is only uprated (increased each year) where there is a legal requirement to do so: the EEA, Gibraltar, Switzerland, and the specific countries that hold a reciprocal social-security agreement with the UK. Everywhere else the payment is frozen at the rate you first received and stays there until you die. This is documented plainly enough in the House of Commons Library briefing on frozen overseas pensions (SN01457) and corroborated across the public record. As of May 2020 about 492,176 people were drawing a frozen UK State Pension abroad. Roughly 84% of them are in Australia, Canada, or New Zealand, which is why the campaign noise has historically come from there and not from Asia.

It is worth being exact about what “frozen” means, because the word does the opposite of clarifying it. Nothing is taken away. There is no deduction, no clawback, no tax. The pensioner simply does not receive the annual increase everyone else receives. In a zero-inflation world that would be survivable. We do not live in one. The freeze is not a cut — it is the slow conversion of an income into a fixed nominal number while everything that number buys moves.

And here is the part that decides the whole thing for a Southeast Asia retiree. Thailand has no agreement. The Philippines does. Thailand is frozen. The Philippines is uprated. Same region, same kind of move, opposite pension outcome, and no relocation brochure for either country puts those two facts on the same page.

The mechanism that does the damage

The increase the frozen pensioner forgoes is the triple lock: each April the State Pension rises by the highest of price inflation, average earnings growth, or 2.5% (House of Commons Library, CBP-7812). It is deliberately generous. That generosity is exactly what makes the freeze expensive, because the frozen pensioner is not missing inflation. They are missing the higher of three numbers, every year — compounding off the prior year’s already-higher base.

Look at what the triple lock actually delivered recently. 2022/23: 3.1%. 2023/24: 10.1%, the largest single uprating in the policy’s history. 2024/25: 8.5%. 2025/26: 4.1%. 2026/27: 4.8%. The frozen pensioner received none of those. Not a reduced version. None.

The retrospective: what it has already cost

Most of this site models the future and says so. This piece does not have to. The freeze has a paper trail, so the artefact here is history, not a forecast, and that is its strength. You cannot argue with a number the government already paid.

Take the cleanest possible case. A person who began drawing the full new State Pension in 2016/17, when it started at £155.65 a week, and went to live in a frozen country. They still receive £155.65 a week today. An identical person who went to an uprating country received every April increase since. Here is the full series of the full new State Pension, the intermediate years reconstructed from the published triple-lock percentages and reconciled exactly to the government-published rates for 2016/17, 2023/24, 2024/25, 2025/26 and 2026/27:

Year	Uprated £/wk	Frozen £/wk	Weekly gap £	Lost that year (×52) £
2016/17	155.65	155.65	0	0
2017/18	159.55	155.65	3.9	203
2018/19	164.35	155.65	8.7	452
2019/20	168.6	155.65	12.95	673
2020/21	175.2	155.65	19.55	1017
2021/22	179.6	155.65	23.95	1245
2022/23	185.15	155.65	29.5	1534
2023/24	203.85	155.65	48.2	2506
2024/25	221.2	155.65	65.55	3409
2025/26	230.25	155.65	74.6	3879
2026/27	241.3	155.65	85.65	4454

Anchor years (2016/17, 2023/24, 2024/25, 2025/26, 2026/27) are government-published rates; intermediate years reconstructed from the published triple-lock percentages. Last column sums to ≈ £19,400 over the decade.

Source: UK published State Pension rates + triple-lock percentages (House of Commons Library CBP-7812, CBP-10403) · checked 2026-05-24

Read the last column down and add it up. The frozen pensioner is, across these ten years, roughly £19,400 worse off than the person who made the identical decision one country to the east. The current-year gap alone is about £4,454, and it does not hold there. It widens every April for the rest of their life, because the uprated figure compounds and theirs is a flat line.

Now sit with the shape of that table rather than the total. The early years look harmless. Four pounds a week. Nine pounds a week. This is precisely why the freeze is not felt at the point of decision and is severe at the point it matters: the damage is back-loaded by construction, arriving in the years the retiree is least able to earn it back. The 10.1% year alone opened a £48-a-week hole in a single April. They did nothing wrong that year. They simply lived in Thailand instead of the Philippines.

Date	£/week	Basis	Note
2016	155.65	sourced	2016/17, system start
2017	159.55	triangulated	—
2018	164.35	triangulated	—
2019	168.6	triangulated	—
2020	175.2	triangulated	—
2021	179.6	triangulated	—
2022	185.15	triangulated	—
2023	203.85	sourced	2023/24, after the 10.1% rise
2024	221.2	sourced	2024/25, after 8.5%
2025	230.25	sourced	2025/26, after 4.1%
2026	241.3	sourced	2026/27, after the 4.8% triple-lock rise

Source: House of Commons Library — State Pension uprating (CBP-7812, CBP-10403) · latest 241.3 £/week (2026) · as of 2026-05-19

Forward, it only bends one way

The retrospective is closed; you cannot argue with rates the government already paid. The only open question is the forward path — and it bends against the frozen pensioner in a way that offends intuition. Project today’s £85.65-a-week gap forward at three uprating paths and the cone does not fan toward relief. It fans toward a larger loss the better the triple lock performs, because the frozen number never moves and the entire gap is the increase they do not get.

Source: Computed forward from the 2026/27 gap (£85.65/wk) at 2.5/3.5/4.5% uprating, frozen rate held at £155.65. A projection, not a forecast · checked 2026-05-24

This sits on top of the ≈£19,400 already gone. Fifteen years out, the frozen pensioner is between roughly £104,000 and £139,000 behind, and the only thing deciding where in that range they land is how well the increase they never receive performs. A pensioner cannot root for a weak triple lock; they are not in it. The policy they have no stake in is the one whose success deepens their hole.

The one variable, isolated

This is the same finding the drawdown model produces from the other direction, and the two pieces are meant to be read together. That model showed that the highest-leverage variable in a 25-year retirement abroad is not investment return but a treaty checkbox, and that the plan fails years earlier in Thailand than the Philippines on otherwise identical inputs. This is that line of the model on its own, costed from the public record. The frozen State Pension is the income that the drawdown model described as being “slowly euthanised by inflation.” Here is the euthanasia, itemised by year, with the government’s own rates as the instrument.

It is also the cleanest illustration of why the geographic cure is a lie in the narrow financial sense. The brochure sells the place. The place is not the variable that breaks you. A rule you never see, attached to the place, is.

The Philippines workaround, and its real price

There is a known piece of folk knowledge here. Thailand-based frozen pensioners are reported to relocate to the Philippines for around six months to have the pension brought back up to the current rate while they are resident there. This is real reported behaviour, not a recommendation, and the mechanics deserve cold precision rather than enthusiasm.

The increase generally applies while you are ordinarily resident in an uprating country. It is not, on the ordinary reading of the rule, a retroactive refund of the frozen years. Move back to a frozen country and the increases stop again from that point. The residency test is fact-specific and is the DWP’s to apply, not a forum’s. Verify your own position with them before treating any of this as settled; this is cited data, not advice.

But cost the manoeuvre honestly, because that is the point of this desk. The “fix” for a frozen pension is to physically uproot a person, frequently an old one, sometimes a frail one, for half of every year or to relocate outright, into a different country, healthcare system, and language, in order to recover an increase that a different address would have granted automatically. That is not a loophole anyone should feel clever about. It is the measure of how much the freeze is worth that people in their seventies rearrange where they are allowed to grow old to escape it.

What would have to be true for the freeze not to matter

The honest section. The freeze is genuinely immaterial to a specific minority, and they are nameable.

It does not matter if the State Pension is a small fraction of your income and the rest is inflation-linked and large: a substantial private or public-service pension carrying its own uprating, drawn in a currency you also spend. It does not matter if you are in an uprating country, the Philippines among them, in which case this entire piece is an argument for one address over another rather than a loss. It matters less the shorter the retirement, which is not a comfort anyone should reach for. And it can be partly neutralised by someone with the means and mobility to maintain residence in an uprating country, which is a strategy available mostly to the people who needed it least.

For everyone else, and that is most people retiring on or near the State Pension, the arithmetic is not an opinion. It is the table above. The triple lock could be weakened later, and note the asymmetry if it is: the uprated pensioner would lose some future growth, while the frozen one is already receiving none of it. Every realistic change to the policy narrows the gap only by harming the better-off case. None of them improve the frozen one. The number on the left of that table moves. The number on the right never has.

Contest the model

The credibility is that you can see where it is wrong for you. Every input, its source, and the direction it moves the result.

• The rate series. Source: UK published rates and the triple-lock percentages (Commons Library CBP-7812, CBP-10403). The anchor years are government figures; the intermediate years are reconstructed from the published annual increases and reconcile exactly to those anchors. Use your own claim year and the absolute gap changes; the shape does not.
• The 52-week basis. A transparency choice. The annualised figures move slightly on a 52.14-week convention. It does not change the order of magnitude or the direction.
• Frozen vs uprated country status. Source: Commons Library SN01457; corroborated publicly. Binary, and the single highest-leverage input. Confirm your own country’s status with the DWP; the list is the government’s and can change.
• The forward projection. The band above runs three uprating paths (a 2.5% floor, a mid 3.5%, a higher 4.5%), explicitly not a forecast. Note the counter-intuitive direction: the more generous the future triple lock, the larger the frozen pensioner’s loss, because the entire gap is the increase they do not get.
• The workaround. Reported behaviour, marked illustrative, residency mechanics fact-specific and DWP-determined. It changes an individual outcome; it does not change the rule.

Rebuild it with your claim year and your country. If the gap over your expected retirement is a number you can absorb without it touching your care years, the freeze is a line item. If it is not, you have learned the most quantifiable thing this site can tell you, and you have learned it while the address is still a choice.

This is cited data and an illustrative model, not financial, tax, or pension advice and not a forecast of any individual’s entitlement. Pension uprating and residency rules are the DWP’s to apply and can change; verify your own position with the DWP and a licensed adviser before acting.