---
title: "Finland's 2026 Budget Framework — Sami Miettinen on Paasipodi #71"
titleOriginal: "Kehysriihi 2026. Mitä jäi käteen? | #paasipodi 71"
description: "Two days after the Orpo government's 2026 spending-limits session, Martin Paasi and Sami Miettinen go through what actually survived: the self-employed pension reform, the rewritten taxation of employee share options, youth employment measures, the household deduction and first-home loans — and then the €17 billion annual deficit underneath it all. Preceded by a quarter-hour on the agent architecture both men are now running."
format: "podcast"
show: "#paasipodi"
episode: "#71"
hosts: ["Martin Paasi"]
date: 2026-04-30
duration: "1:01:45"
language: "fi"
original: https://www.youtube.com/watch?v=u43JkM01nwM
canonical: https://www.neuvottelija.com/media/paasipodi-71-budget-framework-2026/
---

# Finland's 2026 Budget Framework — Sami Miettinen on Paasipodi #71

Recorded two days after the Orpo government's 2026 spending-limits session was published,
this is a line-by-line reading of what was actually decided — with, as usual, a quarter of an
hour of tooling talk first, because both men now spend their nights building.

**The agent stack, briefly.** Miettinen's principal agent **Samantha** runs on OpenClaw on its
own Mac Mini with its own persistent memory structures and skills; his inner circle can query
it directly. Beneath it sits a second AI worker, **Stöbä**, which holds local memory secure
enough for investment-banking material, and a **Hermes** subagent besides. The layered design
is deliberate: secrets stay on hardware he owns, reachable only through SSH and Tailscale. He
is weighing a Mac Studio so that Google's free **Gemma** model can run locally — his framing
being that you fight the American giants by not opening your wallet to them, the €15,000
machine notwithstanding.

Paasi's **Hello Humans** has moved on since episode 66. **Gemini**, the Chinese model **Qwen**
and **Perplexity** now run deep-research bots producing an 8,000-character brief — about ten A4
pages of facts — for every episode, and the debating models must stay inside it; when one
wanders off, the Claude host corrects it against the brief. A Sonnet model orchestrates the
three-model team through prompts. Running costs are around a thousand euros a month.

On data centres, Miettinen breaks with Risto Siilasmaa: build them. Take single data centres
without a strategy around them and nothing comes of it — that much he agrees with — but local
inference capacity is worth having as domestic reserve, and if the hardware ends up
underemployed, Finland gets it cheaply for its own use. He offers to find work for a
ten-billion-euro hall himself if nobody else does.

## The self-employed pension reform (YEL)

Both men put this first, and Paasi calls it the best thing in the whole package. In future a
self-employed person may choose whether their pension contribution is based on **actual
income** or on the notional "work income" figure used today. On the published calculations,
around 40 per cent — the small end — see contributions fall, 40 per cent see no change, and
higher earners pay somewhat more. The reform is not elegant, and the choice-based structure is
not beautiful, but the feedback in the parliamentary group has been strongly positive, including
from people who will pay more.

The substance of the complaint being fixed is the old assessment-based system, under which the
pension cluster picked a notional salary grade for you and issued a distraint-eligible invoice
regardless of whether you were ill or had any cash flow at all. Miettinen knows people who
stopped being entrepreneurs over it, and both treat it as a straightforward barrier to founding
companies — which matters because Finland's largest employing cluster is small and medium-sized
firms, not large ones. Full effect arrives over 28 years; the transition includes 25 per cent
relief tokens for cash-flow crunches. What did not make it is a **funded component**, which
remains under study. Miettinen's dictator's version would have moved ten billion out of the
earnings-pension pot to capitalise it.

## Share option taxation

Miettinen's second pick, and in his view the second-best item in the package. The old rule
taxed options as earned income **at the moment of grant**, on a notional valuation of an
unlisted company — so a person with a €40,000–100,000 salary could face a tax bill of that
order on a paper gain that might never materialise. The predictable result: people declined
options, the instrument fell out of use, and the state's tax revenue fell with it. As Miettinen
puts it, you get less of what you tax more, and here the structural mistake — taxing in advance
— guaranteed there would be nothing at all.

The point of options is recruitment. A five-person startup wanting a genuine specialist away
from a well-paid job needs an incentive: the right to buy shares at a fixed strike price inside
a window several years out. If the strike is €10 and the share is worth €7, the option is worth
nothing; at €20 it is worth ten euros a share.

Under the new treatment, tax falls when the option is exercised or the shares sold. The
understanding Paasi obtained — passed on from a question Jyri Engeström had put on LinkedIn,
and explicitly flagged as not yet nailed down — is that the value from zero up to the strike
price is treated as **earned income**, and the gain from the strike price to market price as
**capital income**. Miettinen's view is that even deferral alone would have been fair, but that
this is the correct split, and he notes for the argument's sake that Finland's 34 per cent
capital income rate is, on the Tax Foundation's European comparison, roughly the third-highest
in Europe. Both regard the old rule as a poison pill in the Finnish startup and unlisted-company
scene, and one of the drivers pushing founders to the United States.

Neither item, they note, was covered much in the press, which concentrated on cuts to
third-sector organisations — and on the corporate tax rate falling to 18 per cent, a decision
taken a year earlier and reported as if it were new.

## Youth employment

The employment voucher rises from €30 million to €50 million; the state hires 500 summer
workers and funds youth unemployment pilots. Miettinen — who describes himself as close to
libertarian on the size of the public sector — supports it, on the grounds that entry-level
work is precisely what AI threatens first, and that this cohort has already been hit twice:
by the pandemic, and by a comprehensive school system whose PISA results have fallen from the
top places to somewhere around thirtieth while Estonia has climbed from where Finland now
sits to where Finland used to be. His practical ask is that every company take on one summer
worker.

## Household deduction, first-home loans, renovation

The household deduction maximum rises from €1,600 to €2,100 and the reimbursement rate from 35
to 40 per cent — the mechanism being that people employ other people. Miettinen adds a warning
from his own experience: work on a housing company's structures does not qualify, and arguing
otherwise earned him a penalty on top of the disallowance, a thing he now checks with an AI
before filing anything.

On **ASP** first-home loans, the borrower's own contribution halves from 10 to 5 per cent, the
state guarantee rises to 95 per cent and the term extends to 40 years. Miettinen is
unenthusiastic — he built his own down payment out of low-cost index funds — but concedes the
instrument works for the risk-averse and that a 40-year term is fine if you know what you are
doing. What he does defend is the habit underneath it: saving ahead is neither capitalist nor
left-wing, it is simply something everyone should do, like brushing your teeth. And put a few
hundred euros into a fund for a child at birth: five thousand euros, left alone, buys that
child an extra average Finnish pension.

€110 million goes to energy renovation grants. Miettinen considers this small change, and — in
a break with much of the press and several economists — is glad the transfer tax was *not*
temporarily zeroed. Two per cent is genuinely low; run the search yourself, he says, before
blaming the transfer tax for a frozen housing market.

## The €17 billion problem

Half a billion in other savings was found, which Miettinen calls a real achievement a year out
from an election even though it is small against a €17 billion annual deficit. His structural
observation — offered explicitly as an observation, not a position — is that without the growth
in health and social care spending, interest costs and defence spending, the whole thing would
balance.

His advice to listeners is unsentimental and aimed squarely at people in a weak position: the
cuts are coming regardless of which party leads the next government, because the money ran out
twenty years ago and the debt is accelerating. The best protection for your own household is to
get free of cash flows somebody else decides, which in his framing means supporting tax cuts
funded by cutting public administration first. He notes he has been unemployed himself, for a
little over a year, and does not think the point is easy.

On the €17 billion's authorship the two spar productively. Paasi stops Miettinen mid-swing over
the €6 billion Uniper loss: whatever one thinks of the outcome, the minister routinely blamed
for it had no ability to influence it — she was not in those rooms, Fortum conducted the
negotiations, and beating one person with that bat indefinitely is unreasonable. Miettinen
takes the opposite tack on principle rather than person: if you want ministers in the
governance structures of state-owned companies, you must also accept the responsibility that
comes with it; if amateurs sit in those structures, you get amateur work, so stop playing at
state ownership. Norway's oil fund, he notes, is run by a former investment banker and is
running full AI integrations. A parliamentary audit committee memo on the Uniper case is coming
to the chamber, which both welcome — Miettinen without expecting it to be gospel.

His own nominee for a multi-billion share of the deficit is not a person but an accounting
convention: for decades the surplus of the pension funds was counted as a state surplus,
masking roughly three billion a year of genuine deficit and letting Finland pretend to be in
balance. It has now been removed, and the gap it was hiding is visible.

He is equally wary of the state treasury, the finance ministry and the pension cluster as
sources: on the interest-rate hedging position, for example, the same people explain the
billions saved earlier without dwelling on the losses booked while unwinding the option
positions. They speak their own book when asked, and if you are not a finance professional
you will not know the right questions to ask.

## Eastern Finland and the third sector

The special programme for eastern Finland continues, with special economic zones under
preparation — a circle drawn on the map inside which tax relief and other measures apply.
Miettinen finds it an interesting experiment.

The larger fight is over the **724 social and health organisations** receiving public funding.
Neither man disputes that the organisations do important work. The question is where the money
goes: to the children and people in distress, or to administration and €150,000 director
salaries at organisations sitting on million-euro portfolios — 724 administrations where one
could do. Fifty million was cut at state level and twenty-five million moved to the wellbeing
services counties; Miettinen would have cut a zero more and pushed the money down to the
municipal level, of which there are 292 and where he sees the same empty offices and
bureaucracy. His constitutional point is the sharper one: if the work matters, fund it from the
budget where parliament decides, rather than routing taxpayers' money to old politicians
appointed as directors with publicly funded assistants beneath them.

## Defence, Ukraine and the exit

Defence spending is the third leg of the deficit, and Miettinen is uneasy about how fast it was
raised, while acknowledging the drones falling into the Gulf of Finland show real gaps, and
that the F-35 will be extremely expensive. Paasi's counterpoint is that growing up inside a
long peace does not mean the peace continues indefinitely. Both note Russia's shutdown of
internet access in the Moscow region and the hunt for VPN users as a self-defeating decision;
both are struck by Ukraine's long-range drone competence against Baltic oil terminals; and both
read the ratification of the €90 billion EU package after the change of government in Hungary
as a genuinely good sign.

Paasi closes with the honest partisan disclosure the episode has been circling: as an
independent he can be critical of the governing parties too, but his verdict on this session is
that real structural reforms were made — and that he awaits a change of government with
apprehension.

Original recording (#paasipodi): https://www.youtube.com/watch?v=u43JkM01nwM

---

Cite as: Sami Miettinen, guest on #paasipodi #71 — Finland's 2026 Budget Framework — Sami Miettinen on Paasipodi #71, 2026-04-30, https://www.youtube.com/watch?v=u43JkM01nwM. Record: https://www.neuvottelija.com/media/paasipodi-71-budget-framework-2026/.