Nordic Economic Policy Review 2024

This article reviews the fiscal policy framework and recent fiscal policy experiences in Denmark. It discusses the issue of how to design and time discretionary fiscal policies and focuses on the difficulties and uncertainties in assessing the business cycle, choice of instruments, shock dependence, challenges of targeting sectoral differences and experiences of unconventional fiscal policy measures. It also covers the role and source of automatic stabilisers and points out that there is an indication that they have become slightly weaker as a consequence of various reforms, which points to a trade-off between incentives and stabilisation/insurance. In recent years, fiscal policy has focused increasingly on medium-term and long-term issues, particularly sustainability, and the article discusses assessments of fiscal sustainability and their implications for policy.

Paper presented at the Nordic Economic Policy Review conference on Fiscal Stabilisers in Nordic Countries, Reykjavik, November 2023. I gratefully acknowledge constructive comments from seminar participants, the discussants Paul Kramp and Søren Hove Ravn and the editors Juha-Pekka Junttila and Jouko Vilmunen.

Keywords: fiscal multipliers; automatic stabilisers; fiscal policy; business cycle

JELcodes: E32, E62, E61

For a number of years, fiscal policy in Denmark has been rule-based, with more weight attached to medium-term and long-term issues and the sustainability of public finances. The degree of activism in aggregate demand management policies and the perception of being able to finetune the business cycle are, thus, different than in the 1970s and 1980s when a more active approach was taken.

Since the recovery from the financial crisis, all standard macroeconomic indicators have been favourable – only interrupted by the COVID-19 pandemic – with economic growth, low unemployment, a surplus on the current account and sound public finances. This is the result of the medium-term focus in economic policy, among other things, and a string of structural reforms supporting high employment rates. The specific features of public finances are summarised in Figure 1, showing that the budget balance in Denmark has been in surplus on average and that gross debt has been reduced (net wealth is positive). The budget position has been systematically better than the EU average, and public sector debt is comparatively low.

Figure 1. Public finances – Denmark and the EU 27, 2000-2022

Note: EMU definitions of deficit and debt, the EU 27 (excluding the UK)
Source: Statistics Denmark

For a small and open economy with a large public sector, particular attention is paid to fiscal policy both in the short term and in the long term. This may explain why Denmark is one of the countries that has consolidated debt, introduced fiscal rules and integrated the issue of fiscal sustainability into policy making.

Although Denmark is an EU member state, it has opted not to join the EMU and pursues a fixed exchange rate vis-a-vis the euro

Denmark pursues a fixed exchange rate policy pegging the Danske krone (DKK) to the euro, formally as part of ERMII, with a +/- 2.25 % band around the central parity (DKK 746.038 per €100).

. This policy was launched in the early 1980s and has since remained an anchor point for economic policy, which enjoys broad political support and high credibility.

Adopting a fixed exchange rate implies a clear division of labour between monetary and fiscal policy since an independent monetary policy is not feasible for a country with liberalised capital movements. The peg essentially implies that the inflation target in the euro-area becomes the implicit inflation target for Denmark, and it is the responsibility of fiscal policy to safeguard the credibility of the peg (see, e.g., Andersen and Chiriaeva, 2007). This means that fiscal policy should ensure a path for competitiveness and economic performance more generally that is consistent with the implicit inflation target implied by the peg, and thereby support the credibility of the peg. Historically, this implication of the peg for fiscal policy was well understood, a point underpinned by the notable policy intervention in 1997 due to a perceived risk that the economy was overheating. At that time, the concern was that the current account surplus would quickly be eroded at the same time as wage growth might pick up and deteriorate competitiveness since unemployment was perceived to have fallen to or below the structural unemployment rate. This led to policy measures aiming at reducing aggregate demand by curbing growth in public and private consumption (via mandatory savings and an increase in the excise duty on private loans). It is noteworthy that the intervention was anticipatory and based on a concern that the economy was on a trajectory inconsistent with the fixed exchange rate policy, and it was perceived that initiative had to be taken before problems grew out of hand. Over the years, macroeconomic developments have, thus, been consistent with the exchange rate peg. The peg has high credibility, as witnessed by the systematically very low – sometimes even negative – interest rate spread to the euro area (Germany).

This paper provides an overview of fiscal policies in Denmark in recent years (although it does not cover the COVID-19 crisis; see Andersen et al. (2022) for a discussion). Section 2 introduces the fiscal policy framework, which has been in effect since 2014. The experiences and challenges that emerge from discretionary fiscal policies are discussed in Section 3, while Section 4 considers the role of the automatic stabilisers and how they have been affected by recent structural reforms. Fiscal sustainability issues and analyses are discussed in Section 5, and Section 6 presents a few concluding remarks.

Fiscal policy planning has become more structured and formalised over the years. Since 1995, it has been customary to prepare long-term plans focusing on fiscal sustainability. This includes medium-term plans (with a horizon of about ten years) typically defining a target for the structural balance in the end-year of the planning horizon, which is then used as the basis for determining the “fiscal space” (råderum) available for new political initiatives either to increase expenditure or to change taxation within the planning horizon. The current medium-term plan has a structural budget balance target of -0.5% of GDP in 2030.

The fiscal framework is defined in a budget act passed in 2012 and effective from 2014, which implements the fiscal rules laid down in the Growth and Stability Pact but goes further by including not only a deficit limit for the structural public balance but also binding and multi-annual (four-year) expenditure ceilings for central government, municipalities, and regions. The Act also defines sanctions if the expenditure ceilings are violated

If municipalities/regions exceed agreed expenditure ceilings, collective and individual sanctions are applied, in the form of a reduction in the block grants they from the State to the municipalities/regions.

.

The Ministry of Finance (2023) summarises the fiscal framework as follows:

While the Budget Act originally specified a deficit limit of 0.5%, it has been changed to 1% of GDP as part of a national compromise on Danish defence and security policy (March 2022). This is in accordance with EU rules since the public debt level is significantly below 60% of GDP.

The governance framework of the Budget Act has been partially deviated from in recent years due to extraordinary circumstances related to the outbreak of COVID-19 (2020–22) and the situation in Ukraine (2022). The presentation of the budget proposal for 2023 marked a return to the normal expenditure management framework.

In addition to the above rules, governments have introduced a “tax freeze”, but it has been interpreted differently over the years. Currently, the interpretation is that any increase in taxes or duties must be accompanied by offsetting reductions in other taxes to ensure that the overall tax share does not increase

Indexation of tax rates and various nominal thresholds to ensure unchanged real values are not comprised by the tax freeze.

. Some excise duties are exempted (e.g., tobacco and nicotine), and revenue from the CO2 tax is returned in full to the respective sectors.

While fiscal activism and the belief in the scope for finetuning the business cycle have been lower in recent years than in the 1970s and 1980s, discretionary changes to fiscal policy remain important. In some cases, they are motivated by the business cycle situation, and in others by political decisions having different motivations. In the latter case it is a question how well these changes are timed to the business cycle situation. A recent example is the discussion about providing help to low-income families to compensate for the high and unexpected inflation in 2022 while avoiding a stimulus that increases inflation.

Discretionary policies are often gauged by considering changes in the cyclically adjusted budget balance. However, it is a summary measure which is estimated with considerable uncertainty and not necessarily a good measure of the aggregate demand effects of a policy change. Denmark has a long tradition of using a fiscal effect metric (finanseffekten) to present the effects of discretionary changes in fiscal policy. This metric measures the total effect on activity (GDP) of a given fiscal policy package as the sum of the changes in the included instruments times their respective multipliers. The fiscal effect is computed for both the immediate effect (one year) and the long term. Both the Ministry of Finance and the Council of Economic Advisors regularly report their assessment of the fiscal effect of implemented and planned discretionary fiscal policy changes.

The computation of the fiscal effects has its roots in Keynesian-type models for the Danish economy with rather detailed modelling of the public sector; the ADAM model used by the Ministry of Finance and the SMEC model used by the Council of Economic Advisors

For information on the ADAM-model, see https://www.dst.dk/da/Statistik/ADAM/Modellen-ADAM. For the SMEC-model, see https://dors.dk/modeller-metoder/smec.

. These models have been refined and updated over the years. While the two models are not too dissimilar, their presence has created an environment of openness and transparency on the quantitative assessments of the effects of fiscal policy. A new model – MAKRO – has recently been developed and is going to be used by the Ministry of Finance. Among the key features of this dynamic model is that it merges short-term effects with long-term structural aspects, see Bonde et al. (2023). One advantage of this is that it provides a more detailed analysis of the fiscal multipliers, their impact and dynamic effects depending on whether they are temporary or permanent, see Røpke et al. (2021).

The fiscal effect, as estimated by the Council of Economic Advisors and the actual GDP growth rate are plotted in Figure 2. Since fiscal policy is planned on the basis of the anticipated business cycle situation, the data should be interpreted with some care. Nevertheless, the figure illustrates several points about discretionary fiscal policies. First, the correlation between output growth and the fiscal effect is about -0.5, suggesting that fiscal policy has, on average, been countercyclical

The simple correlation is not a strict test of the cyclicality of discretionary fiscal policies, since many factors not anticipated when planning the fiscal policy may affect actual GDP.

. Second, the year-to-year variations show the consequence of economic policy changes that have not always been motivated by the business cycle situation but affect aggregate activity, nevertheless. Thirdly, there are episodes where fiscal policy has not been well timed to the business cycle situation. This applies notably to the period preceding the Financial Crisis when the Danish economy experienced a boom-bust pattern partly attributable to fiscal policy being too lean. In the years prior to the crisis, the economy was booming with low unemployment, high growth in private consumption and investments (housing), and fiscal policy was expansionary. However, policy makers were reluctant to tighten fiscal policy

Pedersen and Ravn (2014) use a Taylor-inspired rule to work out the fiscal policy response called for given that monetary policy cannot be deployed as a stabilisation instrument due to the exchange rate peg. They find that developments during this period called for a significant tightening of fiscal policy rather than the expansionary policy which was pursued. They show that this conclusion also holds true based on real-time data, which significantly underestimated the ongoing boom.

. The prime minister, when confronted with calls from several economists to tighten fiscal policy, stated that textbooks had to be rewritten since “this time is different”. The destabilising effects of fiscal policy came from two main sources: a tax freeze and high growth in public spending. The liberal-conservative government introduced the “tax freeze” with the intention of curbing public sector expenditures. The freeze meant that tax rates were not to increase. However, for property taxes and some excise duties, the freeze was defined in terms of nominal tax payments. At a time of rapidly rising house prices, this reduced the effective tax rates, which in turn pushed house prices further up. Growth in public consumption was also high. The average annual growth rate for public consumption was about 2% in the period prior to the Great Recession – primarily driven by increasing health expenditure – while the target for public expenditure growth was about 1%. Fiscal policy was, thus, clearly pro-cyclical during this period. The experience in these years was a motivating factor for the expenditure ceilings later introduced as part of the fiscal policy framework; see above.

Figure 2. Fiscal effects and GDP growth

Note: The fiscal effect measures the one-year effect of fiscal policy on real GDP of fiscal policy changes relative to the previous year.
Source: (a) GDP growth: Statistics Denmark, (b) Fiscal effect: Economic Council, The Danish Economy, various issues.

When discussing discretionary fiscal policies, it is useful to make a distinction between a defensive approach referring to economic policy changes initiated for non-cyclical reasons, but the timing is made dependent on the business cycle situation so as not to be destabilizing, and a more offensive approach actively using discretionary fiscal policy changes to dampen business cycle fluctuations. The former is relevant even if there are reasons to be more sceptical about the ability to pursue an active stabilisation policy. An example of the former is a tax reform in 2004 which due a transition period implied that some tax decreases took effect before other taxes increased, and the timing matched that a temporary expansionary policy coincided with a period with low activity. Another example of problematic timing is a structurally motivated reform shortening the maximal unemployment benefit period from four to two years and a tightening of the employment condition to regain benefit entitlement (from 26 weeks to 52 weeks within the preceding three years). The reform was approved in 2010 to take effect in 2012 when the business cycle situation was expected to have normalised. These expectations were too optimistic (the output and employment gaps turned out to be about -3%), and it became an issue that many unemployed would lose their entitlement to unemployment benefit (and be transferred to social benefits, which are means-tested on a family basis) when implementing the reform in a low activity environment. As a result, several ad-hoc measures were brought in to soften the effects of the reform; see Danish Economic Council (2014) for an account and analysis.

A more offensive use of discretionary fiscal policy to stabilise the economy raises difficult issues besides the political economy aspects, including the information, decision, implementation, and effect lags. Particular challenges pertain to the timing and composition of interventions and their dependence on the nature of the shock(s) driving the business cycle. In particular, a focus on aggregate demand may not be adequate when sectors are affected differently in a specific business cycle; see below.

To find a balance between avoiding finetuning policies, the uncertainty involved in determining the output gap, and yet pursuing an active stabilisation policy, the Economic Council (2007) proposed a guided discretionary rule for discretionary fiscal policy targeting discretionary changes in severe economic situations. Intervention is – according to the rule – only justified if the output gap exceeds (numerically) 1% of GDP. If this condition is met, public consumption and investments should be changed by 1% if the output gap is 1%, corresponding to a fiscal effect of roughly 0.25%. Symmetrically, a similar tightening is called for in the case of a positive output gap of 1%. The Council has more recently re-interpreted the rule to be defined in terms of the employment gap, see Economic Council (2023).

The rule is an interesting starting point for a discussion of discretionary fiscal policy. The following discusses the difficulties in assessing the business cycle situation, choice of instruments, sectoral differences, and experiences of unconventional fiscal policy measures.

Discretionary fiscal policy cannot be changed at short notice, and therefore it has to rely on projections. However, the business cycle situation can change abruptly, as recent developments exemplify. The projections of the employment gap for 2023 by the Economic Council and the government at different projection horizons are shown in Figure 3. The Economic Council’s projections have changed significantly across the different timings of the projections, and the assessments made in the autumn of 2022 had widely different implications for the appropriate discretionary fiscal policy than those made in early 2022 and at the start of 2023. The point here is not that the government had steadier (and more precise) projections than the Economic Council for the specific time period considered; there are other periods when the opposite was the case. The point is to illustrate the large uncertainty in assessing the business cycle situation even within a relatively short forecasting horizon and, thus, to pinpoint the difficulty in planning discretionary fiscal policies. Similarly, there may be large differences between the ex-ante data available when policy is planned and ex-post (revised) data; see, e.g. Cimadomo (2012) and Pedersen and Ravn (2014).

Figure 3. Forecasts of the employment gap for 2023

Note: Spring/autumn refers to the timing of the publication of the Economic Council Reports. The government refers to “Økonomisk Redegørelse” (Economic Survey) published in May and August by the Ministry of Finance/Economic Affairs.

While the rule proposed by the Economic Council is contingent on the size of the output (employment) gap, it is independent of the nature of the shock. This is crucial for the appropriate fiscal policy response. A case in point is the situation in 2022–23, with high unanticipated inflation caused by a global shock. This prompted a tightening of monetary policy, and, given the Danish peg to the euro, the Danish Central Bank raised its rates in line with the ECB. An issue of coordination of fiscal and monetary policy thus arises since the appropriate tightening of fiscal policy clearly depends on the monetary policy responses, an aspect not taken into account by the rule. In other situations, there may, of course, not be an issue of coordination of fiscal and monetary policies given the exchange rate peg, but the example shows one of the problems associated with a shock-independent rule. More generally, the appropriate fiscal response depends on the nature of the shock (demand or supply) and whether it is transitory or permanent. This suggests that a common fiscal policy response to a given output or employment gap is not appropriate.

Discussions about fiscal policy are generally presented in terms of changes in either expenditure or taxes. The design of aggregate demand management policies involves the question of whether to target public or private demand. In the case of expansive policies, targeting private demand mainly involves tax instruments to increase disposable income (e.g. temporary reductions in direct taxes) or induce intertemporal substitution in demand (e.g. temporary VAT reductions). Public demand can be affected via expenditure on both consumption and investments.

However, in practice, it is increasingly clear that the available set of instruments for demand management policies is restricted. Public consumption mainly consists of administration and provision of welfare services, neither of which is ideal for business cycle adjustments, and there are similarly strong smoothing arguments on the taxation side for keeping tax rates steady over time. On the transfer side, adjustments to transfers raise a political economy problem, especially if they are cut. Little room is left for manoeuvre, and public investments are effectively the most flexible instrument despite its limitations. First, the decision lags are long for most public investment projects, which means the instrument can mainly be used asymmetrically as a contractionary measure by postponing a planned project. Secondly, it targets specific sectors (mainly construction), which means it is not generally applicable; see below.

There is also a structural aspect. Traditional macro analyses take a very aggregate approach to the labour market, essentially assuming homogenous labour, which can be reallocated across sectors easily and without costs. The implication is that only the aggregate level of demand matters, and a reduction in one component (say net exports) can be compensated by an increase in another one (say public investments) to mitigate the effects on aggregate activity and employment. This may also be phrased in the way that different components of aggregate demand are perfect substitutes as far as employment is concerned.

This homogeneity assumption may be questioned for several reasons, especially since specific qualification requirements for many jobs tend to make labour non-homogeneous across types of production/sectors, and it follows that the composition of aggregate demand matters. When labour embodies sector-specific knowledge, it is not plausible that the labour can be reallocated across uses at low costs. The costs stem from training (explicit or on-the-job training) and the sluggishness with which workers adapt their reservation demands to changes in their labour market prospects (in particular, if it is associated with a different type of job, lower wage, etc.). Of course, the speed and willingness to adapt depend on wage formation and the design of the social safety net (benefit levels, active labour market policies etc.), including the options it offers for “postponing” the adjustment.

This point is illustrated in Figure 5, showing changes to employment in various sectors during two different periods when the employment rate was falling. Two observations stand out. First, although aggregate employment was falling, there were significant differences, with employment rising in some sectors. Second, the sectors most severely affected were not the same during the two crises, e.g., the construction sector was hardly affected by the first crisis but more severely during the second. Since the sectors differ in the composition of the workforces, it follows that the types of workers affected during the two crises are different. Even this illustration is based on aggregate data not fully displaying the sectoral differences related to the type of labour. Increased specialisation driven by both new technology and globalisation makes the standard macro approach to the labour market more questionable.

Figure 4. Changes to employment by sector; 2001–2003 and 2007–2009.

Note: In the 2002–2004 crisis, total employment fell by 1.4%, in the 2007–2009 crisis by 2%. Employment in persons. Sectors are not weighted by their importance for total employment.
Source: Statistics Denmark

Sectoral adjustment effects like these have important implications for fiscal policy

The effects and design of fiscal policy in the presence of sectoral adjustment costs have not been much researched. One exception is Steigum and Thøgersen (2003). In a full employment model, they allow for the costs of transferring labour from the non-tradeable sector to the tradeable sector. The policy response to a negative private wealth shock comprises both that fiscal policy redistributes from future to current generations by running a deficit (consumers are non-Ricardian) and that demand for non-tradeables is supported in the transition.

. First, even if falling private demand is counteracted by a fiscal stimulus, there will be a transitional phase with excessive unemployment (increasing mismatch problems) due to declining demand in contracting sectors and the sluggish process by which labour is reallocated. It is far from self-evident that a general increase in demand will “lift all boats” in the labour market. The people made redundant may possess other skills than those in demand as a result of a more expansionary fiscal policy. The implication is that shocks may have more persistent effects (including greater wage/unemployment dispersion), and fiscal policy may have a weaker short-term impact due to these supply factors.

Second, it is crucial to determine whether the changes are transitory or permanent. In the case of transitory changes, reallocation of labour may entail inefficiently large adjustment costs due to excessive job turnover. In the case of permanent changes, the issue is more complicated. On the one hand, it is important that the policy does not constrain the necessary structural adjustment process. On the other hand, if unused resources are only sluggishly absorbed in other sectors, there may be an argument for some temporary support even for declining sectors. It is not an easy task to address these issues, since it requires an identification of the sectors facing particular problems and targeting policies accordingly. General measures like tax cuts may not be targeted accurately enough, and sector-specific measures may amount to subsidies, which raises questions both in relation to EU rules and moral hazard problems arising if sectors are bailed out.

In summary, an active use of discretionary fiscal policies to stabilize the business cycle is challenged for several reasons, suggesting that such interventions are only relevant in the event of large shocks and that the automatic stabilisers should be allowed to do their job; see discussion below.

The preceding discussion took a standard approach to the set of fiscal instruments. Therefore, it is interesting to mention some notable examples of unconventional demand management policies targeting private demand in Denmark.

In 1998, a mandatory pension saving scheme (SP) was introduced requiring all those in employment to pay 1% of gross earnings

When launched the scheme was redistributive as contributions were allocated equally among all participants, but in 2002–03 this redistributive element was abolished, and individuals were credited for their personal contributions.

. Originally, this was only meant to apply for a year, but it was made permanent and then suspended in 2004. The motivation behind the scheme was to reduce demand and avoid overheating the economy (a reintroduction of the scheme was actually discussed in 2008 for the same reason). The contributions were accumulated in a funded pension scheme and to be paid out when reaching the statutory retirement age (65 at the time). In an effort to boost private consumption in the wake of the Financial Crisis, individuals were allowed to withdraw their balance in 2009. This is an example of an off-budget demand management policy to increase private consumption and aggregate demand without a negative effect on public finances. Actually, it frontloaded some tax revenue since pension contributions are deductible, but pensions are taxable income. In an analysis of this policy, Kreiner et al. (2019) find that the policy did increase spending for liquidity-constrained households and that the spending propensity was increasing in the tightness of the liquidity constraint. 

Another example of an unconventional demand management policy is the “unfreezing” of holiday pay

In Denmark, a part of wage income (typically 12.5%) is reserved for a holiday allowance paid out during holiday periods. In the past, holiday allowances depended on wage income earned in a previous period (i.e., there was a lag between the accrual of holiday allowances and the period in which the money was paid out). A recent reform synchronised the earnings and holiday periods, and to avoid a double pay-out of holiday allowance, one part was frozen until retirement age. In response to the COVID-19 crisis, it was decided to allow individuals to request a pay-out of the frozen holiday allowances in two rounds (autumn 2020 and early 2021).

during the COVID-19 crisis, which simultaneously both improved the disposable income of households and tax revenues since holiday allowances are taxable income. In the autumn of 2020, holiday pay corresponding to DKK 31 billion (1.4% of GDP) was paid out, followed in early 2020 by DKK 22 billion (1% of GDP). This had a considerable impact on households’ disposable income.

It is unclear whether such policies can be used again in the future. The example of the holiday payment arose by chance due to the reform of the holiday pay system (and motivated by the experience with the release of the mandatory pension savings). Generally, it is not advisable to allow a short-term factor to influence the design of pension systems. The SP pension arrangement was motivated by the experience that mandatory pension schemes reduce consumption and therefore have a contractionary effect, and an expansionary when released. But it is not clear whether this could be repeated. As a response to the unanticipated high inflation in 2022–2023, some transfer payments have been temporarily increased, e.g. to pensioners, partly to mitigate the consequences of the time lag in the indexation of transfers (basically, all transfers are indexed by wage increases). These measures can be interpreted as examples of unconventional fiscal policy, although they were not motivated as a measure to support aggregate demand. More general use of cash transfers during deep crises raises political economy issues in ensuring that such changes are temporary and targeted to groups with a high marginal propensity to consume.

Automatic stabilisers are widely appreciated as the rule-based part of fiscal policies not suffering from the same challenges as discretionary fiscal policies. Therefore, calls are regularly made to strengthen the automatic stabilisers (and thereby also to reduce the use of discretionary fiscal policies). In the wake of both the Financial Crisis and the COVID-19 pandemic, calls have been made to strengthen automatic stabilisers, see e.g., IMF (2023). The fact that monetary policy is constrained by the zero-lower bound on the interest rate has also been put forward as an argument for the need to strengthen automatic stabilisers because it is difficult to time discretionary fiscal policy in relation to business cycle fluctuations; see, e.g., Blanchard and Summers (2020).

The automatic budget response or stabilizer is a summary concept for the automatic response of public sector revenue and expenditure to changes in the level of economic activity (the business cycle). They are the net outcome of other policy choices, which then produce these automatic responses, including taxation, implying that tax revenue changes when income changes and the social safety net when social expenditures changes, when (un)employment changes, and so on.

Automatic stabilizers work primarily by stabilizing private consumption – a key component of aggregate demand (and thereby also indirectly private investments). This is generally considered an important and valuable part of fiscal policy. Automatic stabilisers cushion individuals’ disposable incomes and, therefore, serve an insurance purpose because they have a direct and positive welfare effect for risk-averse agents. They also contribute to stabilization of the aggregate economy via their stabilising effect on disposable income and hence private consumption and aggregate demand; see, e.g., Van der Noord (2000). Moreover, they tend to mute the consequences of economic crises on income inequality; see, e.g., Domeij and Flodén (2010), Dolls et al. (2011c), and OECD (2014).

Automatic stabilisers are rule-based automatic responses to changes in the business cycle. Hence, they do not require up-to-date information on the state of the economy, and they do not require any discretionary policy actions to work. Therefore, they generally work more swiftly, counter-cyclically, and more targeted than discretionary policies, being subject to information, decision, and implementation lags (see above).

A necessary condition for automatic stabilisers to work is fiscal space making room for the implied budget variations. Maintaining symmetry across the business cycle is important. Budget surpluses (and, therefore, consolidation) during upturns create room for budget deficits and automatic stabilisers to work during downturns. If such fiscal space is missing, discretionary tightening of fiscal policies may be called for during a recession, which counteracts the effects of the automatic stabilisers. There are many examples of countries offsetting automatic stabilisers by discretionary policy changes due to lack of fiscal space. Price et al. (2015) find that in nearly half of the OECD countries, automatic fiscal easing was accompanied by discretionary tightening for half of the period 1980–2018. Prudent fiscal policy in good times is, therefore, an important precondition for automatic stabilisers to perform their countercyclical role in bad times. In Denmark, it has been an important argument in favour of fiscal rules that they safeguard the space to allow the automatic stabilisers to do their job.

In assessing automatic stabilisers, two aspects are particularly important: the nature of the shock (demand or supply) and its persistence (temporary or permanent). In general, the optimal policy response depends on the nature of the shock, while automatic stabilisers “average” across types of shock. Automatic stabilisers work best to cushion demand-driven business cycles. They do not distinguish between temporary and permanent shocks. Since it is possible to diversify temporary but not permanent shocks, this is important. The effects of aggregate shocks are reflected in the budget balance accumulated over time if shocks are persistent, as seen during, for example, the Financial Crisis. The implication is that automatic stabilisers can never be set to “autopilot”. If shocks are persistent, close monitoring and intervention are needed to avoid that public debt comes on an unsustainable trajectory.

While automatic stabilisers are widely discussed and praised, there is no well-defined or commonly accepted way to measure their strength. The literature features assessments taking a microeconomic approach based on microsimulation models capturing the details of tax and transfer schemes, macroeconomic models allowing for various behavioural and equilibrium responses, or a statistical approach assessing the sensitivity of the public budget to variations in output

This is measured by the semi-elasticity of the budget balance to a change in the output gap; that is, by how many percentage points the budget deficit/surplus (measured relative to GDP) changes following a 1% increase in GDP.

. For a brief overview and references; see, e.g., Mohl et al. (2019). The first two approaches are resource-demanding and model-specific, which is why the statistical approach is widely used. It is relatively simple and allows accessible time series for public revenue and expenditure to be used; see, e.g., van der Noord (2000), Debrun et al. (2008) and Price et al. (2015). The issue with this approach is that the estimated relationship is shock-dependent (depending on the sample period used in the estimation), and it captures budget responses which are not necessarily closely related to the aggregate demand effects being important from a stabilisation perspective. For instance, variations in the revenue from corporate taxation are important from a budget perspective but less so for the short-term aggregate demand effects. Despite the widespread use of this metric, it is thus not clear that it appropriately measures the links contributing to stabilize the incomes of households and thus aggregate consumption and demand; see Andersen (2016) and Maravalle and Rawdanowicz (2020) for a discussion. It should be noted that the methods referred to above find that Denmark is among the countries with the strongest automatic stabilisers, and its level has been relatively stable over time; see, e.g. Price et al. (2015) and Maravalle and Rawdanowicz (2020).

The following takes a different and more theory guided approach, focusing on the part being both most important for aggregate demand and directly related to policy instruments (see Appendix). The starting point is the basic source of the automatic stabilisation of private consumption and thus aggregate demand that arises via the effects on disposable income originating from changes either on the intensive margin due to changes in wages and/or hours worked or on the extensive margin from changes in employment due to business cycle fluctuations. For the former, the marginal tax rate is the relevant metric of how changes in income affect disposable income and, hence, the potential for private consumption. On the extensive margin, the relevant metric is the participation tax measuring how disposable income is affected by a shift between employment and non-employment. The participation tax depends both on the tax rate and the generosity of the transfers received when out of work (e.g., unemployment benefit or social assistance). In practice, the participation tax varies between groups in the labour market, and in general, the budget effect is, therefore, the sum of changes in employment in different groups multiplied by their respective participation taxes.

Over a sequence of years, policy initiatives have focused on the strengthening of work incentives to work via both changes in the taxation system and the design of the social safety net. This has contributed to reduce the number of people in the working-age population receiving public transfers and increased the structural employment rate. Figures 5 and 6 show the developments in the marginal tax rates on earned income and the participating tax (both identified as the key transmission links in the Appendix), and in both cases, there is a downward trend

If a tax cut is accompanied by a broadening of the tax base, the net effect on the stabilisation of disposable income is ambiguous because the latter implies that a broad income base is covered by the stabilisation mechanism. Broadening tax bases has been more important for corporate taxation (for example) than household taxes.

. Thus, these changes have weakened the automatic stabilisation of household disposable incomes. This is not to imply that the policy changes are sub-optimal, but it does point to the difficult of strengthening work incentives without weakening the automatic stabilisers. There is a trade-off between efficiency and insurance (stabilisation).

Figure 5. Marginal income tax rates for earned income, 1993–2023

Note: The income tax system used to have three tiers, but the middle tier was abolished with effect from 2009. The taxes include the earned income tax credit (beskæftigelsesfradrag) introduced in 2004, but not the additional tax rebate for singles with children. Marginal taxes are computed using the average municipal tax and do not include the church tax.
Source: Ministry of Taxation, www.skm.dk

Figure 6. Participation taxes for individuals eligible for unemployment insurance or social assistance

Note: The participation tax for single people without children (important for social security).
Source: www.oecd-ilibrary.org

The preceding discussion makes it clear that the automatic stabilisers are the net outcome of decisions made on the taxation system and social safety net. It is a straightforward implication that the more extended the welfare state, the higher the tax rate and social transfers and, therefore, the more sensitive the budget is to changes in private employment.

The fact that the automatic stabilisers are the net outcome of other policy choices and not the outcome of a separate policy decision is often overlooked when calling for them to be strengthened. This interrelationship also brings out an important trade-off between incentives and insurance (stabilisation); see, e.g., Gruber (1997), Knieser and Ziliak (2002), and Andersen (2016). As an example, a high level of participation taxes does, on the one hand, strengthen the automatic stabilisers and the ability to cushion shocks, but on the other hand, it weakens the incentive structure with detrimental effects on structural employment.

Summing up, automatic stabilisers play an important macro role. They are rule-based and are the net outcome of policy decisions on the tax structure, social safety net, etc. The automatic stabilisers are, therefore, not the result of macro considerations but other decisions in other policy areas. As a consequence, reforms to strengthen work incentives have tended to weaken the automatic stabilisers, and this raises the question of finding other ways of strengthening them without jeopardizing work incentives, such as, for example business cycle dependent unemployment insurance; see discussion in Andersen (2023).

The consequences of ageing and the implications for public finances have taken centre stage in economic policy discussions in recent years. Assessments of fiscal sustainability are regularly made and provide important input to economic policy discussions. The key metric is the sustainability indicator giving the permanent change in the primary budget balance (as a % of GDP) needed for current policies to be consistent with the intertemporal budget constraint for the government.

During the first phase, the issue was to understand the nature of the demographic changes and to quantify the challenges and need for reform. This led to a string of reforms, including (in particular) increases in and subsequent indexation of the statutory retirement age to the development in longevity. These reforms ensure that the conditions for fiscal sustainability are met; see Ministry of Finance (2023), Economic Council (2023), Hansen et al. (2023). During the second phase, the focus has been on ensuring that policies remain consistent with fiscal sustainability, and this is implemented via the setting of short-term targets for fiscal policy; see above on the fiscal policy framework.

Since 2015, both the government and the Economic Council have assessed fiscal policy to be sustainable, but the sustainability indicator has varied over time, and the assessments made (with the sustainability indicator varying between -0.2 and -1.8; that is, leaving some room for spending hikes/tax cuts), see Economic Council (2023). These different assessments also point to the underlying uncertainty in the sustainability metric, which is no surprise given its forward-looking nature and the fact that it depends on a long list of assumptions.

The specific aspect of the Danish case is that the underlying budget profile displays a U-pattern (in policy debates known as “the hammock”); see Figure 7. The starting point is favourable (including low gross debt and a net wealth position), followed by a sequence of years with deficits and then systematic surpluses (note the deficit on the total balance is smaller than on the primary balance due to the positive net wealth). In present value terms, the future surpluses are sufficient to cover the series of deficits (taking into account the initial net wealth position).

The main reason for this pattern is that although retirement ages are linked to longevity

Statutory ages in the pension system (for public pensions and early retirement, as well as age limits for payments from pension schemes) are established by law and, thus, regulated at the political level. Recent reforms—the 2006 Welfare Reform and the 2011 Retirement Reform—have increased the statutory retirement ages in steps from age 60 years to 64 years for early retirement (2023), for the public pension from 65 years to 67 years (2022), and also shortened the early retirement period from five years to three. The second element in the reforms is indexation of the early retirement age and pension age to the development in life-expectancy at the age of 60 in order to target the expected pension period to 14.5 years (17.5 including early retirement) in the long term (currently about 18.5/23.5 years). Parliament decides every 5^th year with a 15 year lead the statutory retirement age, hence in 2020 it was decided that the statutory retirement age in 2035 will be 69, and in 2025 it is up for approval that it is going to be 70 years in 2040. There is a speed limit such that the statutory retirement age can only be increased by a maximum of one year every 5^th year.

, the scheme has a speed limit implying that it takes several decades (according to current population projections) to reach the target for the indexation, which is an expected retirement period of 14.5 years for all cohorts. Hence, the speed limit implies that the expected retirement period for a sequence of cohorts exceeds this target. The current retirement age is 67, and it has been decided that it will be 68 from 2030 and 69 from 2035. Since future increases in the retirement age must be approved by parliament, there is an inherent risk in the underlying budget profile. The next decision is coming up in 2025, and according to the indexation rules, the retirement age should be increased to 70 with effect from 2040. If there are no further increases beyond 70, the primary balance will remain in deficit, and fiscal policy will not meet the sustainability requirement, see Kommissionen om tilbagetrækning og nedslidning (Danish Pension Commission, 2022).

Figure 7. Projected developments in public finances, Denmark, 2020–2080

Source: Council of Economic Advisors (2023).

While the sustainability indicator usefully summarises a great deal of information in a single metric and is relatively easy to communicate, it also suffers from the problem that it conceals the uncertainty inherent in such projections (which increases with the length of the forecast horizon). While this can be circumvented to some extent by presenting – as is usually done – robustness analyses, the problem remains that the indicator is often interpreted too literally, i.e. without taking into account the uncertainty in the assessment.

Moreover, the indicator takes outset in the question of whether existing policies are fiscally sustainable; it is a feasibility test and not an optimality test. Plausibly policies and behaviours will be changed going forward, including higher expenditure on health and lower working hours, all of which would tend to have a detrimental effect on fiscal sustainability. Nonetheless, sustainability analyses have served their purpose and have been instrumental in giving greater weight to medium-term and long-term aspects of fiscal policy making, which has, in turn, contributed to minimise the role of myopia and zig-zag tendencies in policies.

The major shift in fiscal policy making in recent years reflects both less optimism about the potential to finetune business cycles and focus more on medium-term and long-term issues. This is explained by both the poor track record of active demand management policies during the 1970s and 1980s and the looming challenges of an ageing population. Institutionally, these changes are reflected in fiscal policy frameworks that aim to boost the continuity, consistency and credibility of fiscal policy making, and more predictable policies may in themselves be conducive to economic developments.

Discretionary fiscal policies remain relevant, although the challenges of timing such interventions to the business cycle developments are generally recognised. There is a consensus that such interventions should be reserved for “large shocks”, although there is less agreement on how to define “large”. In contrast, there is agreement on the virtues of automatic stabilisers, and calls are often made to strengthen them since they are rule-based and respond automatically to changes in the business cycle situation. However, this is easier said than done since the automatic stabilisers are the net outcome of decisions on the tax system, the social safety net, etc. A side effect of recent reforms designed to strengthen incentives to work has, therefore, been to weaken the automatic stabilisers. This points to a fundamental trade-off between incentives and insurance/stabilisation.

Andersen, T.M. (2016). Automatic stabilizers – The Intersection of Labour Market and Fiscal Policies, IZA Journal of European Labour Studies, 5(11).

Andersen, T.M. (2023). Fiscal Policy Challenges, Nordregio project on “Economic policy beyond the pandemic in the Nordic countries”.

Andersen, T.M. & Chiriaeva, J. (2007). Exchange Rate Pegs, Fiscal Policy and Credibility, Open Economies Review, 18(1), 53-76.

Andersen, T.M., Holden, S., & Honkapohja, S. (2022). Economic developments and policies during the COVID-19 crisis – Nordic experiences, Nordic Economic Policy Review, 13-50.

Blanchard, O. & Summers, L. (2020). Automatic Stabilizers in a Low-Rate Environment, AEA Papers and Proceedings, 110, 125-130.

Bonde, M., Ejarque, J., Høegh, G., Partsch, E., Stephensen, P., & Vasi, T., (2023). MAKRO - Model Documentation: A Handbook for Using and Understanding the MAKRO Model. DREAM, Copenhagen.

Cimadomo, J. (2012). Fiscal Policy in Real Time, The Scandinavian Journal of Economics, 114(2), 440-465.

Dolls, M., Fuest, C., & Peichl, A. (2011). Automatic Stabilizers, Economic Crisis and Income Distribution in Europe, Research in Labor Economics, 32, 227-255.

Debrun, X., Pisani-Ferry, J. & Sapir, A. (2008). Government Size and Output Volatility: Should We Forsake Automatic Stabilizers, IMF Working Paper wp/08/122.

Domeij, D. & Flodén, M., (2010). Inequality Trends in Sweden 1978-2004, Review of Economic Dynamics, 13(1), 179-208.

Economic Council (2004). The Danish Economy, Copenhagen.

Economic Council (2007). The Danish Economy – Spring, Copenhagen.

Economic Council (2014). The Danish Economy – Autumn, Copenhagen.

Economic Council (2023). The Danish Economy – Autumn, Copenhagen.

Gruber, J. (1997). The Consumption Smoothing Benefits of Unemployment Insurance, American Economic Review, 87(1), 192-205.

Hansen, J. Z., Andersen, M., & Dalgaard, T.N. (2023). Langsigtet økonomisk fremskrivning 2022 – vurdering af den finanspolitiske holdbarhed, DREAM.

IMF (2023). Fiscal Monitor, Washington.

Knieser, T.J. & Ziliak, J.P. (2002). Tax Reform and Automatic Stabilization, American Economic Review, 92(3), 590–621.

Kommissionen om tilbagetrækning og nedslidning (2022). Fremtidssikring af et stærkt pensionssystem, Copenhagen.

Kreiner, C. T., Lassen, D.D., & Leth-Petersen, S. (2019). Liquidity Constraint Tightness and Consumer Responses to Fiscal Stimulus Policy, American Economic Journal: Economic Policy, 11(1), 351-379.

Maravalle, A. & Rawdanowicz, L. (2020). How Effective are Automatic Fiscal Stabilisers in the OECD Countries?, Economics Department Working Paper, 1275, OECD.

Ministry of Finance (2023). Denmark’s Convergence Programme 2023, Copenhagen.

Mohl, P., Mourre, G., & Stovicek, K. (2019).  Automatic Fiscal Stabilisers in the EU: Size and Effectiveness, European Economy - Economic Briefs 045, European Commission.

OECD (2014). Society at a Glance, OECD, Paris.

Pedersen, J. & Ravn, S.H. (2014). A Taylor Rule for Fiscal Policy in a Fixed Exchange Rate Regime, Danmarks Nationalbank Working Paper no. 90.

Price, R.W.R, Dang, T. T. & Botev, J. (2015). Adjusting Fiscal Balances for the Business Cycle: New Tax and Expenditure Elasticity Estimates for OECD Countries, OECD Economics Department Working Paper 1275.

Røpke, J., Høegh, G., & Kronborg, A. (2021). Finanspolitiske multiplikatorer i MAKRO, En sammenligning med andre modeller og empiri, Baggrundsnotat, DREAM, Copenhagen.

Steigum, E. & Thøgersen, Ø. (2003). Borrow and Adjust: Fiscal Policy and Sectoral Adjustment in an Open Economy, International Economic Review, 44(2), 699-724.

Van der Noord, P. (2000). The Size and Role of Automatic Stabilizers in the 1990s and Beyond, OECD Working Paper, 230.

Van der Noord, P., Girouard, N., & André, C. (2006). Social Safety Nets and Structural Adjustment, OECD Working Paper 517.

To identify the key source of the part of automatic stabilisers important for private consumption, consider a stylised setting in which total household disposable income is given as

(1) 

where t denotes the tax rate, w_p the wage rate in the private sector, w_g18 the wage rate in the public sector, b the level of social transfers to non-employed

It is assumed that all transfer payments are taxable income (as is the case in some countries), but this is not crucial to the arguments.

, L_p the employment level in the private sector, L_g the employment level in the public sector, and N the number of recipients of social transfer payments (not in employment). At the business cycle frequency, public wages and employment are constant.

The business cycle is associated with changes in the private sector: i) wages or hours worked (intensive margin responses) and ii) employment (extensive margin responses). For the intensive margin response, it follows straightforwardly that

(2)

This captures the Domar-Musgrave effect that income taxation reduces the variability of disposable income. The higher the tax rate, the less disposable income varies relative to variations in the taxable income.

Turning to the extensive margin changes, the total population (P) is made up of employed in the private and public sectors and the non-employed

This assumes that everybody out of a job is entitled to the transfer, which is a reasonable approximation for the Nordic countries

(N), . Variations in private employment (for given ) are mirrored in changes in the number of non-employed, . Hence, the effect of changes in private employment on disposable income is

(3)

which can be rewritten as

(4)

where  is the so-called “participation tax” for the individual when transitioning between work and non-work, and is the replacement rate of the transfers. To see this, note that the difference between income when working and non-working is

This makes the participation tax crucial to the response of disposable income to changes in employment. It captures the essence of the automatic stabilising effect from employment variations arising from the taxation of income and the social safety net providing income transfers to the non-employed.

Finally, to see the relation to the budget responses, the public sector primary budget balance (B) is in this stylized setting given as

(5) 

where T denotes other sources of tax revenue (exogenous), and G other public expenditure (exogenous). Note that the tax rate should be interpreted broadly as capturing both income and consumption taxes.

In this simple formulation, there is no distinction between income and consumption. Similarly, profit income is disregarded (taken to be exogenous).

Observe also that in Denmark – and in most OECD countries – more than 90% of tax revenue accrues from the direct and indirect taxation of labour income, and about 2/3 of public consumption is wage expenditure; s, hence the above captures the main effects on the budget.

The budget effect of a change in private employment (for given public employment ) is

or

(6)

The total budget effect of a transition of one single individual from non-work to work in the private sector is , i.e. the sum of the tax paid and the after-tax value of the social transfer. The transition from work to non-work has a double effect on the budget, both the direct loss of tax revenue from reduced private income and the extra expenditure on social transfers

In the case where there is only a change in wages in the private sector, the budget effect arises solely from the tax side and the automatic stabiliser is smaller.

.