Introduction
The COVID-19 pandemic that started in 2020 resulted in economic disruption across a large part of the world. Many governments duly responded. One example was the extension of short-time working (often called Kurzarbeit) schemes, which avoided layoffs and paid more generous compensation than did many regular unemployment benefit programmes. In some countries, unemployment benefit periods were extended and/or job retention and special business loan schemes were established. Across the world, a variety of special/temporary measures, some more targeted than others, were initiated to protect what were considered the more vulnerable members of society (IMF, 2020; ESPN, 2021; Casey and Mayhew, Reference Casey and Mayhew2023; ILO, 2023).
One important case of such special measures was that of granting people access to pension savings that normally would have been available to them only once they had retired. In a wide range of countries, participation in such retirement savings accounts had been made mandatory, and savings were granted a privileged tax status – contributions were tax exempt, as were the returns on investment. Although early access to pension savings occurred in emerging market economies, where welfare states are less developed, it also occurred in advanced economies with much more developed welfare states. Moreover, although the pandemic eased by the end of 2021, in some of the countries early withdrawal opportunities have been repeated or, at least, proposals have been made for their repetition. This article explores cases of premature pension withdrawals to meet income shortfalls, the political context in which the relevant programmes were introduced, and their potential consequences. Its ability to compare across a variety of examples, rather than to concentrate upon only one, is one of its advantages.
There might be short-term benefits for governments taking such measures as well as for the people who take advantage of them. This could well hold particularly if other components of social protection are underdeveloped – social assistance programmes are limited and/or of job furlough schemes are absent. Nevertheless, people tend to withdraw pension fund money everywhere these policies are initiated. Withdrawals can also be very efficiently implemented within days in cases where an urgent response is required. Although such programmes could contribute to inflation in the short term, and so have implications for interest rates that impact on jobs further down the line, this has been less of a problem than some predicted. However, the withdrawals observed do seem to have had distributional consequences. In most cases, they tend to benefit the better-off in the short term. They also reduce pension income and potentially increase reliance on indirect taxation in the longer term. More importantly, they undermine the concept of a pension fund, which is to save for retirement. Overall, the politics of premature withdrawals have a populist flavour as well as contributing to growing uncertainties with respect to retirement income.
The rest of this article is organised as follows. Section 2 embeds the paper within the discussion of pension reforms. Section 3 considers some cases of early access to pension savings: it does so by looking, first, at examples in emerging market economies and, second, at the less frequently mentioned examples in advanced economies. Section 4 reviews the political context of the measures discussed. Section 5 considers the outcomes of the early access to pensions, including responses to, and criticism of, the policy. Finally, Section 6 presents the conclusions.
A new role for funded pension accounts
Pensions are a key, visible dimension of welfare states. Since their emergence in the late 19th century in Bismarck’s Prussia, and even more so in the period following the end of the Second World War, they usually took the shape of pay-as-you-go systems. Public systems were often designed to produce benefits reflecting contributions or earnings made when in employment, but many also contained a solidarity component involving a minimum, or even a maximum, pension. They also often made provision for time involuntarily out of the labour market due to sickness, unemployment, education, or military service (Barr, Reference Barr2020). However, after the 1980s, another paradigm became influential – one centred on workers paying into individual accounts managed by bodies specialising in financial investment. Under such systems, it was not usual for any contributions to be made for time spent outside formal employment (Orenstein, Reference Orenstein2008). The money saved in such schemes can be accessed at retirement. Annuitisation arrangements exist in some countries, but even where they do, people reaching the required age are often able to take their savings as a lump sum, which they can, then, either reinvest and/or draw down in stages.
This form of privatisation became, at least for a period of time, the dominant pension policy promoted by major global actors such as the World Bank, the International Monetary Fund, and other international organisations (Béland and Orenstein, Reference Béland and Orenstein2013; Lindert, Reference Lindert2021). It was believed to be a necessary step for reducing the burden on public finance of demographic ageing. For developing economies, privatisation was also seen as a way of mobilising finance for investment in growth. For the ex-socialist countries of eastern Europe, privatisation was also seen as establishing a retirement system more appropriate for them in their new-found status as market economies.
The emergence of the COVID-19 pandemic led to a new role being given to privatised pensions – as an access to money in cases of income shortfalls. In welfare states, the redirection of policies in terms of function or purpose is not new; such cases may result from changes in the environment, limited knowledge of the designers, and spaces created for new ideas of actors, and so forth (Streeck and Thelen, Reference Streeck and Thelen2005). A well-known example is the use of the public pension system in Germany. Initially, it was intended to provide insurance for disability at work. It morphed into an old age pension system, simply because those working above 70 (later 65) were deemed to be, by definition, work-disabled. By the early 1970s, the system was used to help deal with enterprise restructuring and overall economic downturns – rather than becoming unemployed (and, thereby, adding to the “statistics”), older workers could be early retired. Indeed, it was argued that they had had their turn, and they should make way for young people (Heinelt, Reference Heinelt1991; Schmähl, Reference Schmähl, Rein and Schmähl2003).
This article seeks to answer the two questions: whether pension savings withdrawals are an appropriate response to “temporary” income shortfalls; and what happens to the pension systems that rely on saving funds should “withdrawals” become a more normal policy?
Examples of withdrawal programmes
More than 30 countries allowed some form of early withdrawal from funded accounts, but the conditions for these are strictly controlled (Madeira, Reference Madeira2024). What happened following the onset of the COVID-19 pandemic was both novel in what was permitted but also remarkable in terms of the extent to which the opportunities created were taken up. Utilising information from fresh academic literature, reports of national institutions and international organisations, as well as the media, this article analyses the most noteworthy examples of withdrawals during the period from spring 2020 onwards. They come both from the emerging market economies of Chile, Peru, Kosovo, and Malaysia, and from the advanced economies of Australia, the United States, and Iceland.Footnote 1
The case of emerging market economies
Although Chile and Peru are next-door neighbours, they are also “not so far apart” from Kosovo. Each are prominent examples of the “neoliberal” paradigm of individualised savings. Kosovo’s pension system, although introduced nearly 30 years later than that of Chile, was very explicitly modelled upon the latter’s system. The Malaysian pension system is rather different. It was developed out of a colonial era “provident fund” system, but it too is based upon individual accounts, with the amount saved depending upon fund performance. However, because it is a provident fund, money can also be taken out before retirement to meet designated special needs.Footnote 2
Social security provisions in Chile, Peru, Kosovo, and Malaysia – including unemployment benefits and social assistance schemes – are much less encompassing or even non-existent compared to most developed countries. When COVID struck, there existed no permanent scheme for short-time working, even if Chile and Peru both introduced a temporary “furlough” scheme – something the United Kingdom also did – and Kosovo introduced both a job retention scheme and a temporary unemployment scheme (IMF, 2020).
Chile
The Chilean economy was hit very hard by the pandemic, and an intensive lockdown strategy was pursued – see Table 1. In June 2020, the government legislated to allow people with pension accounts to take up to 10 per cent from their savings pots to provide them with a temporary uplift to their incomes (Carrera and Angelaki, Reference Carrera and Angelaki2022). The then president, Sebastián Piñera, opposed the move. The governor of the central bank described the effects of the measures as being non-linear, cumulative and increasingly irreversible, and argued that the associated costs would rise, such that withdrawals would put pressure on inflation, and the risk of disruptions in the financial market would increase (FIAP, 2021).Footnote 3 So, too, did the Chilean pension funds, which regarded the move as contrary to the retirement scheme’s purpose.
Key indicators for countries with premature withdrawals
Sources: Informality, ILO and national data; Blavatnik index at https://www.bsg.ox.ac.uk/research/covid-19-government-response-tracker; Inflation, IMF.
However, congressional approval was almost universal. Moreover, what had been a one-off exercise was repeated twice more. By the conclusion of the third round in mid-2021, 39 per cent of account holders had made at least one withdrawal. Some 3.8mn of about 9.7mn accounts held by the population had been reduced to zero. Total assets in pension funds were reduced by $50bn, or by at least a quarter (Madeira, Reference Madeira2022).
In spring 2022, Piñera was succeeded by a new, “radical” president, Gabriel Boric. As a member of congress, he had supported the withdrawal plans. However, once he was in office, he acted to prevent a fourth round. Various alternatives were proposed – including allowing access only under strict conditions for specific purposes and providing opportunities for people to make additional contributions to help repair their savings pots (Reuters, 2022). Subsequently, and as the president’s popularity went on the wane, these proposals came to nought (Financial Times, 2022a). Nonetheless, a related proposal was resurrected in spring 2024 by a large fraction of the Chilean congress. This time, the argument was that, since the funds belonged to the savers, they should be allowed to access them for their own investment purposes, including for home purchase or construction. Up to 15 per cent of the account should be available for withdrawal (Carrillo, Reference Carrillo2024).
Peru
Peru’s pension reform was influenced by developments in Chile but was implemented only in 1993. In its character, the reform was very similar to that of Chile, although the transition to the new system was not mandatory (Queisser, Reference Queisser1998). The government first permitted withdrawals in April 2020 – this being even earlier than was done in Chile. Up to 25 per cent of holdings could be taken in three parts and subject to 30-day gaps. A maximum of $3,553 was set. At the same time, workers who had contributed in the past but had been unemployed for at least six months were able to take out a one-off payment of about $600. When the then president, Martin Vizcarra, first proposed a one-off withdrawal, he described it as an “emergency measure.” However, congress took the idea and ran with it. As one commentator said, “the government opened this small door, then of course the politicians got the idea to expand it” (Bernal, Reference Bernal2020). And expand it, they did. The April 2020 withdrawal provision was repeated in June of the same year, while in November of that year, a further provision was made for contributors who had been unemployed for at least 12 months.
Not that these moves were made without opposition. The ministry of finance and the central bank declared that a massive sale of pension fund assets would affect the value of the country’s sovereign bonds, thus making future borrowing more expensive. They also pointed out that such sales could have a negative impact upon the domestic stock exchange, while if the assets that were sold were foreign, this would result in a massive influx of dollars, severely affecting the exchange rate (Bernal, op cit.). As in Chile, there were criticisms that the policy damaged the legitimacy of the pension system – a system that had taken the past 30 years to build up (Americas Quarterly, 2020).
The November proposal was also challenged by the executive as being unconstitutional on the grounds that savings in pension funds were not merely private property, to which savers should be entitled, but they also constituted a collective savings fund, the purpose of which was to ensure the payment of pensions in the future. The constitutional court upheld this view (Tribunal Constitucional, 2021). Not that its judgement had any effect. A further law of March 2021 allowed people aged 40 and over and who had not contributed in the past 4 years to take out the entire contents of their accounts. Last, and in the context of the war in Ukraine and its implications for consumer prices, contributors were able to withdraw a further amount – to a maximum of $4,890. On top of this, those who were employed were able to temporarily suspend their contributions. This made for a total of six withdrawal packages over a period in which Peru also had five presidents.Footnote 4
The amount withdrawn up to and including the fifth package has been estimated to be about 37 per cent of total pension assets (OECD, 2021; FitchRatings, 2022). Some 65 per cent of the eight million pension fund members made some withdrawal, and some 45 per cent of these were people who were currently working. According to some estimates, half of the funds would be completely emptied by the time the fifth programme had run its course (el Commercio, 2021). By the time the sixth withdrawal had been completed, withdrawals had totalled $23bn – the equivalent of 9.4 per cent of GDP (Peru Support Group, 2024).
In spring 2024, and against the instincts of President Dian Boluarte and the ministry of economy and finance, the Peruvian congress pushed through a seventh withdrawal. It justified this on rather similar grounds to those used in Chile – namely that savers could do better than the pension funds with their own money. It has been suggested that as much as a further $7bn could be taken out of retirement accounts (Reuters, 2024).
Kosovo
Kosovo embraced pension reform only in 2002. As a breakaway from Serbia, Kosovo’s creation as a state was one of the final stages of the disintegration of what had once been Yugoslavia. After the breakup of Yugoslavia, each of the new states enacted reforms, some more radical than others. In Kosovo, this process went further than elsewhere. The system it adopted mirrored that of Chile almost in its entirety, and it did so very explicitly (Gubbels et al., Reference Gubbels, Snelbecker and Zezulin2007; Casey, Reference Casey2022).
Kosovo followed Chile’s suit with its own withdrawal scheme. In December 2020, around the time of the Chilean second round, the centre-right government legislated to allow savers to take 10 per cent of their savings pots (Official Gazette of the Republic of Kosovo, 2020).Footnote 5 Again, parliament’s support was across the board, with only one, left-wing party opposing it. Some two thirds of savers took advantage of the provision and 10 per cent of the pension assets were withdrawn (KPST, 2021).
Once a withdrawal process had been initiated, people wanted more. In early summer 2022, legislation was proposed to permit a withdrawal of up to 30 per cent of remaining funds (Kosovo Assembly, 2022). The leading party in the new government, which was now of the left, declared itself against the plan, but most of the main opposition parties backed it. The trade unions said that 90 per cent of their members wanted the proposal to be passed (Koha, 2022). A people’s petition was also submitted to the parliament. The IMF, however, cautioned strongly against it (IMF, 2022). Ultimately, the proposal failed.
One difference to Chile was that the government in Kosovo agreed to reimburse the pension funds for any individual withdrawal up to €999. Although partly protecting pension savings, this proposal generated a potential increase in public expenditure. Kosovo had, historically, sought to maintain a fiscal deficit not exceeding two per cent of GDP. Total withdrawals amounted to some €200mn, and the state ended up taking on reimbursement liabilities of just over half of this amount – close to 1.4 per cent of what was the country’s 2020 GDP. Such reimbursement started in 2023, with the plan of prioritising payments to people near retirement. The programme is supposed to be completed in 2028.
Malaysia
As a “provident fund,” the Malaysian pension system had permitted members to make withdrawals from part of their accounts to meet designated health costs, to finance housing, to pay for education, or to undertake the Haj (EPF, 2022). With the onset of the COVID-19 lockdowns in April 2020, a first withdrawal programme was announced. This permitted limited monthly payments to be taken by people who were still below normal pension age. The withdrawals had to be from that part of the account that was accessible for early access under provident fund arrangements – not from that part of the account that was reserved for pay-out on retirement. This programme was scheduled to end in the spring of 2021.
However, it was followed by three further initiatives, with the first starting in spring 2021 and the last starting in spring 2022. Under these initiatives, money could be drawn down from the retirement savings part of accounts. Moreover, higher amounts could be taken out at any one time or over the prescribed period. By the conclusion of the final initiative, some $33bn had been withdrawn. This accounted for some 15 per cent of the total amount in the members’ accounts.
With respect to who was making use of the facility, it seems as if it was primarily people who already had relatively low amounts in their accounts. This had its own implications. It was already understood that most people reaching retirement age had assets sufficient only to meet 4–5 years of retirement (Smart Investor, 2020; The Star, 2024). The use of withdrawals exacerbated what was already perceived as a challenge to the system. Indeed, it was for this reason that Prime Minister Anwar Ibrahim, warning that 81 per cent of pension contributors would not have enough savings to allow them to live above the poverty line once they had ceased working, decided in spring 2023 that there would be no further withdrawal programmes (The Star, 2023). There were also suggestions that arrangements for additional voluntary contributions into the system could be made so accounts could be replenished (KWSP, 2021).
Malaysia went further in spring 2024 when another opportunity for early withdrawal was established. In order to entice self-employed people and those who were not mandated to join the existing provident fund, a new component was built into the existing system. This allowed immediate withdrawals, regardless of reason, from the non-age-pension part of the account. A pent-up demand for such withdrawals had been identified. Estimates have been made that between $5-12bn (or 1.2–2.9 per cent of GDP) could be taken out of savings accounts in the short to medium term (Yeap, Reference Yeap2024).
The case of advanced economies
Although the most cited examples of pension withdrawals come from emerging market countries, similar programmes were also adopted in Australia, the United States, and Iceland – albeit on a smaller scale. Since the early 1990s, Australia has had a mandatory “superannuation” scheme that dominates its old age pension system. In the United States, voluntary, employer-sponsored savings plans have largely replaced corporate defined-benefit plans for those employees who have supplementary pension coverage. Iceland also uses savings plans as a component of its retirement system. Unlike in the United States, these plans are both mandatory and voluntary, and they cover a substantial share of the working population.
Australia
Australia’s mandatory superannuation system has built up assets of nearly 130 per cent of GDP. Although the country was not hit hard by the pandemic in so far as infection levels and deaths were low, it was effectively shut off from the rest of the world by banning inward travel. This led to its own disruption – see Table 1. A response to that was a one-off measure, announced in spring 2020, that permitted two opportunities for withdrawal from superannuation funds – one to take place in the current tax year that ended at the end of June, and the second in the tax year immediately succeeding it. Withdrawals were limited to $4,720 on each occasion, and they were restricted to people who were unemployed, or eligible for welfare payments, or who had had their working time cut by at least 20 per cent. Previously, early access had been restricted only to “exceptional circumstances” or after at least six months of unemployment. Total withdrawals were estimated to be the equivalent of about one per cent of superannuation assets, so the scale of the measure was much smaller than in Chile, Peru, Kosovo, and Malaysia. The average total withdrawal was about $3,780.
Some 15 per cent of the savers took advantage of the measure, and of these, four out of 10 participated in both rounds (Bateman et al., Reference Bateman, Dobrescu, Liu, Newell and Thorp2023). According to one retrospective study, nearly 1mn accounts were either fully or nearly completely cleared, although given that people might well have had more than one account, this figure has to be treated with some caution.Footnote 6
In addition, the Australian government relaxed the amount that could be drawn down by people who had already retired. This had been restricted to five per cent in any 1 year. Temporarily, the limit was lifted to 10 per cent. Moreover, the measure was subsequently extended for the next tax year.
The United States
A much-overlooked component of the United States’ CARES (Coronavirus Aid, Relief, and Economic Security) Act of 2020 was that it temporarily granted tax-free access to up to $100,000 from employer-sponsored 401(k) individual pension accounts, from defined-benefit accounts, and from individual (not employer-sponsored) retirement accounts. Access was dependent upon the consent of the specific scheme; thus, not all people could participate automatically. As in Australia, under normal circumstances, early access was permitted only in cases of termination of employment, disability, or death.
One study indicated that nearly three-quarters of plan sponsors did permit participants to access retirement funds if needed. Of those who were entitled to, nearly six per cent did so, with 69 per cent making one withdrawal and the remainder two. The average withdrawal was $15,700, and the median was $6,500. However, as nearly one-third of participants who initiated a withdrawal did so more than once, and could do so from more than one plan, the average withdrawal per person was approximately $24,600, with a median of $13,300 (Vanguard, 2021). On some estimates, up to $48bn was withdrawn, but this constituted only a small proportion of the country’s GDP (Derby et al., Reference Derby, Goodman, Mackie, Mortenson, Mitchell, Sabelhaus and Utkus2024).
Less is known about the distribution of withdrawals. However, a separate study that examined withdrawals from pension funds in the United States during the “great recession” of 2008–2010 (so before the special provisions introduced during the COVID crisis) indicated that it was lower-income families who were much more likely to experience the sorts of shocks that lead to withdrawals. They were also more likely to make a withdrawal when they experienced those shocks (Argento et al., Reference Argento, Bryant and Sabelhaus2015).
Iceland
In Iceland, special withdrawal provisions were made available for those participating in the third pillar of the country’s pension system. Although this pillar is voluntary, collective agreements mean that most employees are members of third-pillar schemes. Pillar III schemes can be accessed early for expenditures on first-time house purchases, or at age 60 (IMF, 2023). Special withdrawal provisions were opened in spring 2021 – so relatively late after the onset of the pandemic.Footnote 7 They were initially scheduled to last until the end of 2021, but they were extended until the end of June 2023. The amount permitted for withdrawals was set quite high, with a maximum payment per month of some $6,000 (at a time when median full-time wages were about $5,300) and a total cap of about $91,000. Over the period to April 2023, about $280mn was withdrawn – the equivalent of some four per cent of total fund assets (Bank of Iceland, 2023).
Summary
A sketch of the outcomes of the withdawal programmes is provided in Table 2. This covers the number of rounds of withdrawal, the number of people who participated, the impact on pension savings that followed, and the size of the injection these savings had relative to GDP.
Details of premature withdrawal programmes (2020–2023)
Source: Authors’ calculations based upon data provided. See text for individual countries, plus national accounts.
Note: Excludes seventh round of withdrawals in Peru (April 2024) and new EP3 account in Malaysia (April 2024). Also excludes proposal for further withdrawals in Chile.
The politics of the withdrawal process
Institutions representing more orthodox economic interests, be they central banks, organisations representing pension funds or international financial institutions (IFIs), were always critical. In the case of Peru, the Constitutional Court was critical, too. It was certainly possible that some countries might have been able to respond better – for example, by borrowing to finance temporary social protection measures. Three emerging market countries – Chile, Peru, and Kosovo – had low fiscal debts and had the opportunity to undertake more conventional responses that would not have shocked financial markets.Footnote 8 Moreover, IFIs and others seemed to be prepared to lend.
In Chile, Peru, Kosovo, and Malaysia, initial withdrawals were pushed forward by centre or centre-right parties. In these countries, such parties might have been expected to protect pension funds. However, the introduction of pension withdrawals was, effectively, a populist response by all of the four governments – populist in the sense that the withdrawals were easy to implement in response to the perceived needs of the public, whether or not the consequences were well thought through.Footnote 9 Once the first moves had been made, most political parties, often trade unions, and even citizens wanted the programmes to be repeated. The fourth withdrawal that was proposed in Chile was justified as being an opportunity for individuals to invest better and more cheaply than government-mandated funds (Carrillo, Reference Carrillo2024). Similar arguments were used to justify the seventh Peruvian withdrawal provision. Here, the anti-competitive behaviour of the four government-approved funds was decried, with the implication that, left alone, individuals could achieve superior performance (Peru Support Group, 2024).
It is also notable that some declared left-wingers rejected the measures, especially once they found themselves with executive responsibilities – for example, Boric in Chile, and the Self-Determination Movement (Lëvizja Vetëvendosje) in Kosovo. In both cases, the result was that groupings that might have normally criticised the market-oriented system of personal accounts instead found themselves defending it from interference. Equally, it was a more centre-left Alliance of Hope (or Pakatan Harapan) government of Malaysia that deemed it necessary to put a stop to initiatives introduced by its predecessor, even if that same government subsequently introduced its own opportunities for a form of early withdrawal.
A degree of populism could also be distinguished in Australia. A week before the general election of May 2022, the centre-right prime minister, Scott Morrison, whose government was trailing in the polls, attempted to woo voters by pledging to allow first-time homebuyers to raid their retirement savings for a house deposit. This proposal was duly condemned by pension funds and by economists – both arguing it would fuel a surge in prices in what was an already overheated housing market (Financial Times, 2022b). The centre-left opposition party won the election, and the proposal went no further.
Outcomes of the withdrawals
The withdrawal initiatives potentially served their objective as temporary smoothers of consumption. Fears about their impact on inflation, which were expressed by many critics, as well as fears about their impact on capital markets and exchange rates, proved less of an issue than might have been suggested – perhaps because some monetary authorities took necessary corrective action. In the case of Chile, the IMF observed that market resilience could be ascribed to the quick policy responses by the central bank and to the depth of the country’s financial markets (IMF, 2021). It is to be noted that such fears were mainly expressed with respect to emerging market economies. When the IMF was commenting on advanced economies under its Article IV process, it made no mention of their withdrawal programmes.Footnote 10 Of course, one further reason for this was that the measures were one-off and/or also much smaller.
Perverse consequences
It was, in fact, in the advanced economies that some perverse consequences have been observed. In the case of Australia, it appears to be that “withdrawers were not on average suffering a temporary wage shortfall” (Hamilton et al., Reference Hamilton, Liu, Miranda-Pinto and Sainsbury2023). Moreover, the same study indicated that the largest single item for which withdrawals were used was gambling (Hamilton et al., Reference Hamilton, Liu, Miranda-Pinto and Sainsbury2023). Other studies from that country indicate withdrawals were used for discretionary spending as much for as anything else (Warren, Reference Warren2021; Bateman et al., Reference Bateman, Dobrescu, Liu, Newell and Thorp2023). Similar findings have been reported for the United States (Berger, Reference Berger2020; Holmes, Reference Holmes2021).
In emerging markets economies, the pattern was somewhat different. In the case of Chile and Peru, it seems as if it were often those with the lowest accumulated savings and/or the lowest earnings who took money out (Olivera and Valderrama, Reference Olivera and Valderrama2022; Fuentes et al., Reference Fuentes, Mitchell and Villatoro2023). Less information is available for Malaysia, although one qualitative study indicates that the attraction of the scheme was greatest for low-income groups who had relatively immediate needs (Jiton and Ibrahim, Reference Jiton and Ibrahim2024; The Star, 2023). These people were more often ethnic Malays (Bumiputera), and this was an additional cause for concern (Lim, Reference Lim2023). In the case of Kosovo, all but the groups with the lowest levels of savings in their accounts exercised the option to take money out, and this was irrespective of whether or not these savers were likely to benefit from later replenishment.Footnote 11
This is not to say that other critical comments have not been made. As most observers have pointed out, as a form of relief, the withdrawal measures tended to be regressive. They were so because, particularly in emerging markets, they cover only a minority of the population – those who work or had worked in the formal labour market – see Table 1. Yet the regressivity argument also applies to some of the advanced countries. Certainly, the Australian provision allowing higher rates of drawdown by those already retired has been criticised as benefitting the better-off retirees, who could then use part of their savings for bequests (Davis, Reference Davis2021).
A second round of regressivity arises from the fact that depletion of savings accounts results in lower pensions. According to an early OECD (2020) study, a 10 per cent withdrawal in 1 year could result in retirement income being reduced by between two and nine per cent, with those taking money late in their working lives suffering the most. Of course, in some cases, those with lower pensions might be compensated by being able to benefit from a guaranteed minimum pension. This might well occur in Chile, where earlier and contemporaneous reforms have resulted in the introduction of a pension supposed to compensate for those whose income from the funded system prove too low. According to one estimate, up to 92 per cent of losses might, in fact, end up by being replaced in this way (Madeira, Reference Madeira2024). In the case of Kosovo, compensation was to be realised by the replenishment of accounts. Yet the financing, either of minimum pensions or of replenishment payments, is, ultimately, covered by general taxation. In both Chile and Kosovo, the major source of tax income comes from levies on consumption.Footnote 12 Yet such taxes tend to be regressive, so it is the less well-off who are compensating those who were able to build up pensions at all. In countries where there is either no, or only an inadequate, minimum pension, family members might find themselves under pressure to top up the pensions of their parents. This, too, means that the withdrawal measures could have redistributive consequences – by “re-traditionalising” families.
Apart from the macroeconomic, fiscal, or distributional consequences of withdrawal programmes, they are guilty of a fundamental deficiency in so far as they undermine the validity of savings-based pension systems. This charge was made frequently by central players in emerging markets, and it was repeated by external bodies, such as the IMF. It was also made with respect to at least one developed economy – Australia – where the provision for enhanced drawdowns was criticised on such grounds (Investment Magazine, 2022). Indeed, if pension funds do become a source of income in times of temporary hardship, then effectively they cease to be pension funds at all. Instead, they become provident funds.
With respect to such funds, the widely admired case of Singapore is worth of consideration. The funds it brought in contributed to an extensive programme of public housing. However, the fact that people can access their accounts for a variety of reasons, and particularly to buy their homes back from the state, has led to those who do reach retirement finding themselves in the situation of being “asset rich but income poor.” This holds even if many older people, because they own their own homes, no longer have to make payments on rent (WBG, 2019), and it suggests that provident funds might not always be successful in assuring adequacy well-being in retirement. It is to be noted that, in Malaysia, where although a third fund was added to its pension system, enabling more, and less restricted, withdrawals than before, the proportion of contributions dedicated to the retirement fund, was simultaneously increased.Footnote 13
Possible alternative policies
In the case of the COVID pandemic, many governments were faced with a social risk, the dimensions of which fell outside the perceptions both of most social policy commentators and outside the expectations of many policymakers. That risk was not insurable, except by the state, and its duration was uncertain, too. Conventional social policy tools were likely to be insufficient – this being one reason why governments introduced or extended existing short-time working schemes. Governments also introduced a plethora of other measures, including aids to business, extensions to unemployment benefits, and top-ups to social assistance payments.
Some of these measures might have been superior in their efficiency to the granting of temporary withdrawals from pension funds. In some cases, however, governments might not have had the infrastructure or capacity to develop alternatives.Footnote 14 If this were so, a measure such as granting pension fund withdrawals might have been all that was immediately available.
This leaves open the question of what might be done in the future. It is a question that is relevant not only for the emerging market economies that were the principal users of the policy described in this article, but also for advanced economies. Moreover, it is a question that remains relevant even if the COVID pandemic has abated. The likelihood of future pandemics remains present.Footnote 15 There have been some opportunities to assess the COVID-induced measures that were introduced in 2020 and 2021. It is likely that some will not be repeated, even if they are also modified. Some measures taken were highly wasteful, and assertions that money was badly spent, and even fraudulently claimed, resound in many of the advanced economies. In Australia, the testing of eligibility criteria under its withdrawal scheme has been argued to have been seriously deficient.Footnote 16 To the extent that there were failures in that country, it is plausible to think there were failures elsewhere.
One proposal that did attract limited attention before the COVID crisis, and is thus completely unrelated to it, has been to attach a supplement to individual pension savings accounts which would allow for non-retirement-related withdrawals. This scheme is often described as a “sidecar” scheme. It would enable additional contributions to be made into an account, over and above the level deemed necessary to generate an appropriate pension, with the withdrawal of surplus money being subject to certain contingencies.Footnote 17 Adding such a supplement would assist decision-making insofar as it recognised that individuals often engage in a “mental accounting process” whereby savings are not necessarily completely fungible.Footnote 18 However, even were sidecar schemes to be more widely introduced, they might still suffer from the charge of regressivity, since it would be the better-off who were more likely to participate.Footnote 19 Moreover, they would do little for emerging market economies, where only a fraction of the labour force would be saving in pension funds in any case.
Effective social policy is difficult to achieve, and in emerging market economies, it is even more difficult. Capacity building remains an important priority, even if it should be recognised that this takes time. In the interim, expect there to be more less-than-perfect policy responses.
Conclusion
In response to the COVID-19 pandemic, a number of governments in both emerging and advanced economies opted to provide premature access to pension savings. In some cases, up to 40 per cent of people emptied their savings accounts. Although the money taken had fewer inflationary implications than some had feared, and it might even have helped to sustain economic activity and so employment, the programmes initiated tended to be regressive in their character – doing little for those on smaller incomes or working outside the formal labour market. Furthermore, they threatened to reduce pension adequacy over the longer term, and they might well have shifted the burden of retirement costs in ways that were less than desirable. As social protection policies, they can be criticised both as populist and as less than fully thought through.
