Election posters by Isamaa: “Your money is burning in the 2nd pension pillar” & “We will extinguish” Source: ERR

 

Article by Heidi Reinson, RTU Riga Business School’s Baltic Finance Centre; University of Tartu

Latvians and Estonians have always enjoyed friendly rivalry. Latvians joke about Estonians being slow, and Estonians tease Latvians for always being one step behind. But the tables have turned. Five years ago, Estonia rushed into a sweeping pension reform, allowing people to withdraw their entire accumulated 2nd pillar retirement savings at any age. Now Lithuania is introducing similar changes, and Latvia is debating whether to follow. Fortunately, Latvia has the advantage of hindsight.

Estonia’s 2021 reform has become a unique natural experiment in pension policy. In a new policy paper published by the Baltic Finance Center at RTU Riga Business School, I examine what happened over these five years after the reform, as well as the political and economic dynamics that led to it. The aim is not only to assess its impact, but also to clarify key design details and consequences that are often misunderstood or insufficiently discussed in Latvia. The findings offer both warnings and practical lessons for neighbours considering similar paths.

The most radical “pension freedom” reform in Europe and beyond

Five years ago, Estonia carried out one of the most radical pension reforms in Europe. The mandatory funded 2nd pillar became fully voluntary, allowing anyone, at any age, to withdraw their entire accumulated 2nd pillar retirement savings in one go. Those who withdrew were barred from rejoining the 2nd pillar for a period of 10 years. No age limit, no partial option, just “all or nothing”.

It is one thing to decide to reform a pension system to make it more flexible, but it is quite another to think through the exact design. It is important for Latvia to understand that the design of the Estonian reform made it one of the most radical “pension freedom” reforms ever introduced. In comparison, the UK introduced its “pension freedoms” in 2015, allowing partial withdrawals only from age 55. Australia permits early access under hardship rules. But Estonia went further than anyone.

The reform was driven by a political campaign centred on “freedom.” The historic inflexibility of the system, combined with low visible benefits, had created fertile ground for a political narrative promising to put “your money in your hands.”

The media campaign was not subtle (see the picture below), and highly visible slogans of “freedom” won out over technocratic warnings about future poverty.

The reform sparked fierce debate, with rhetoric reaching extremes. The same policy was simultaneously described as both a historic injustice and a historic liberation. Then-opposition leader Kaja Kallas called it “one of the greatest disgraces in Estonia’s recent political history”, warning of rising poverty in senior age, while former Hansapank CEO Indrek Neivelt compared it to “the abolition of serfdom/slavery two hundred years ago”.

The party and the hangover

When the first withdrawal window opened in September 2021, 150,000 people, or nearly 20% of eligible savers, cashed out immediately. As a result, €1.3 billion flowed into household bank accounts in a single month, equivalent to roughly 4.5% of GDP. By the end of 2025, 37% of eligible participants, or more than 250,000 people, had exited. In total, over €2.3 billion has been withdrawn.

The short-term effects were visible. Retail sales rose sharply, home improvement spending accelerated, and banks recorded a surge in deposits. Based on data from the Estonian Central Bank, around 30% of withdrawn funds were used to repay consumer loans, reducing household debt burdens. Approximately half of the money remained in bank deposits a year later, while about 15% flowed directly into consumption. But the relief was temporary.

The hangover that followed was real. The consumption boom fuelled inflation, adding an estimated 1–2 percentage points to the cost of living just as energy prices began to soar.

A less discussed fact is that the state also benefited fiscally. Withdrawals generated roughly €450 million in income tax revenue. Moreover, when individuals exited the 2nd pillar, the total mandatory contributions did not decline. Rather, the 4% share of social tax that had previously been directed to the 2nd pillar funded pension now flowed into the pay-as-you-go system. With roughly 250,000 exits and average wages around €2,000 per month, this shift represents an estimated €200–250 million per year redirected from private accounts to the current state budget.

The two Estonias

The interesting story lies not in how much was withdrawn but in who withdrew and who did not.

Exit rates were highest among individuals with lower education, lower income, and larger families. Roughly half of 2nd pillar participants with primary education withdrew, compared with about one quarter of university graduates. Parents with three or more children exited at rates approaching 60%. Thirteen percent of first-wave withdrawers had no official income in the year prior to reform. For many, this wasn’t a calculated investment decision. It was an unexpected opportunity for extra cash, driven by necessity or deep mistrust in the state and financial institutions.

The average withdrawal was around €8,000 or roughly seven to eight monthly net wages. For financially constrained families, it provided short-term relief, but put into long-term perspective, it is very concerning. Estonia’s public pension replacement rate currently stands at just 37.8% of average earnings, far below the OECD average of 63% and is due to decrease in the future.

And the divergence did not stop at the withdrawal decisions. In effect, the Estonian society is now splitting into two very different retirement paths. On one side, higher earners who stayed in the system are using the new flexibility introduced in 2025 to increase their contributions to 4% or 6% (nearly 100,000 people have done so). They are also more likely to save in the 3rd pillar. On the other, lower earners who exited lost out on years of compound growth, and will have to rely on a shrinking 1st pillar state pension.

The complexity of decisions introduced by the reform makes matters worse. Before the reform, people only had to make one 2nd pillar decision by choosing a pension fund, but after the reform, the range of options to consider and trade-off exploded. People now must decide whether to stay in the 2nd pillar, exit, increase contributions, change funds, or manage investments themselves. For financially literate households, flexibility can be an advantage. For vulnerable households, complexity can amplify risk, hence the benefits of the freedom in pension systems are not evenly distributed.

The missing debate on the effectiveness of guardrails

In many countries, the proportionality of pension freedoms is defended through “guardrails” meant to minimize rash decisions. But do they work in practice?

To prevent impulsive decisions, the Estonian government initially introduced two guardrails: income tax on withdrawals and a 10-year re-entry ban. The tax could have had some deterrent effect, particularly for higher-balance savers. The re-entry ban, however, functioned less as a behavioural safeguard and more as a rigid penalty. For a family struggling to pay bills today, ten years is abstract and meaningless. Instead of acting as a deterrent, this rule has become a trap. Those who regret their decision are now locked out, unable to rebuild their savings even if they want to. The government is currently discussing reducing the no-entry ban to five years.

Another important design feature is the so-called “cooling off” period. The (up to) 5-month waiting time is a friction that can prevent impulse withdrawals. This aligns with emerging behavioural research on beneficial frictions, deliberate design elements that slow decision-making in contexts where impulsive choices may be harmful. Estonia’s processing delay appears to serve this function, creating space for reconsideration. As a contrasting anecdotal example, in South Africa a newly introduced early withdrawal since 2024 can be executed with one click in an app. It is remarkably easy to withdraw pension savings, potentially leading to impulsive decisions.

What our neighbours should learn

Latvia has the luxury of learning from Estonia’s mistakes. This time, being a little slower might be the smartest thing you do.

Discussions about the need for greater flexibility in the 2nd pillar are inevitable. However, they must be far more nuanced and need to consider the pros and cons of a much broader range of options, as the choice is not between the status quo and a fully open door to cash out. There are tweaks that can be made that expand freedom while still protecting retirement security.

Among its peers, the Estonian reform was not a “standard” path but rather an extreme “all-or-nothing” outlier. Many countries have chosen middle-ground approaches instead. For example, New Zealand permits early access only in defined hardship cases or for first-home purchases. The Netherlands is introducing a 10% lump-sum option at retirement, within a system built on much higher contribution rates. Even the UK’s widely cited “pension freedoms” reform retained a minimum withdrawal age of 55.

Also, if one is to justify a change by the fact that guardrails have been introduced, there needs to be real evidence that they will work. Estonia’s experience shows that the “all-or-nothing” withdrawal rule pushes people towards full cashouts when partial access might have preserved some savings. Moreover, the re-entry ban does not prevent impulsive decisions but keeps people out of the 2nd pillar scheme for long periods of time.

Overall, one has to acknowledge that such reforms shift the responsibility of saving for retirement from the state to the individual. But decades of behavioural research tell us that people are notoriously bad at making decisions for their future selves. Handing them unlimited freedom without effective safeguards is a gamble, with retirement security as the stake.