6. The economic and social impact of low fertility in Norway
As most other OECD countries, Norway’s population is rapidly ageing. With an increasing share of the population reaching pension age as well as more and more requiring long-term care, the country will face substantial demographic challenges over the coming decades. Whether different trajectories of future fertility rates will have a noticeable influence on these pressures is not necessarily clear from the outset. Using the OECD Long-Term Growth Model, this chapter projects demographic, economic, fiscal, and social outcomes under different fertility scenarios in Norway and selected other OECD countries until 2060.
Across many OECD countries, the demographic trends over recent decades have led to high degrees of population ageing. While in 2020 about 18% of the population on average in OECD countries were aged 65 years or above, only 13% were so in 2000 and 11% in 1980. This is both a result of a continued increase in life-expectancy across most OECD countries, as well as a direct consequence of the fall in the fertility rates that many OECD countries have experienced since the second half of the 20th century (OECD, 2017[1]).
A continued ageing of societies will have substantial implications for economic growth, productivity and the sustainability of public finances. While fewer people will enter the workforce, an increasing share of the population will reach retirement age and leave the labour market. All else equal, this would put countries on a path of limited productive capacity and economic growth. At the same time, demographic changes towards an ever-older society critically raise fiscal pressures as government expenditure increases under potentially decreasing public revenue when the workforces shrink and productive capacity stalls (Rouzet et al., 2019[2]; Crowe et al., 2022[3]).
Continued declines in fertility would clearly increase such pressures over the mid- to long-run, yet it is not clear whether a stabilisation of fertility rates – or even a rebounding to previously seen levels – would be sufficient to lead to demographic, economic and fiscal sustainability in the future. To answer this question, this chapter looks at different scenarios on the development of fertility rates in Norway and other selected countries and analyses how different fertility rates could affect the demographic development in the years to come. Using the OECD Long-Term Growth Model (Guillemette and Turner, 2018[4]), the chapter further projects the economic and fiscal consequences of such different fertility trajectories.
The effects of future fertility rates on the economic and fiscal future of Norway have previously been examined by the Norwegian Ministry of Finance (2021[5]) and Statistics Norway (2019[6]), finding initially slightly decreased fiscal pressure under lower fertility scenarios, but overall reduced room for fiscal manoeuvre irrespective of future fertility rates. This implies that when aiming to sustain growth, public finances and societal well-being, aiming for higher fertility rates may not be the right approach. Instead, more focus should be put on supporting Norwegians to have the numbers of children they wish to have, as well as prolonging working lives, increasing long-term investment in private pension savings and steering a growing share of the workforce into long-term care profession while improving productivity in all sectors of the economy – including long-term care and health services (Goldin, 2022[7]; Skirbekk, 2022[8]; Gietel-Basten, Rotkirch and Sobotka, 2022[9]; Ministry of Health and Welfare, 2023[10]). The following sections add an international view, putting projections for Norway in the context of fertility developments in Norway’s neighbours and selected other OECD countries.
6.1.1. Main findings
While Norway is expected to age substantially at baseline, a lower-than-expected fertility rate – converging to 0.5 below baseline TFR – would increase such pressures. The population share of the elderly (65+) is expected to increase from 18% to 26% between 2020 and 2060. Under a low fertility scenario, this could even increase to 29%, while high fertility – converging to 0.5 above baseline TFR – would limit the size of the elderly population to 24%. The growing share of the elderly population means that relatively fewer individuals of working age will support them. Between 2020 and 2060, the old-age dependency ratio is set to increase from 27% to 44% at baseline in Norway, which is the second largest relative increase projected across the considered countries (after Iceland).
For all fertility scenarios, the size of the Norwegian labour force is set to increase between 2020 and 2060, just as in Iceland and Sweden. However, while the baseline scenario projects a 10% increase between 2020 and 2060, low fertility may limit such increases to 1%. In Denmark, Finland, France and Germany, low fertility could reduce the size of the labour force by 10-20% by 2060.
By reducing the population base that shares economic output, lower than baseline fertility in Norway would increase average annual potential GDP per capita growth between 2020 and 2060 by 0.09 percentage points relative to baseline. Similar dynamics can be observed for all considered countries. Beyond the projection horizon of this chapter, low fertility is nevertheless projected to have negative implications for average annual potential GDP per capita growth.
While these findings need to be considered with caution, Norwegian public primary expenditure – government spending excluding any interest payments – is set to increase by about 6 percentage points of GDP in 2060 relative to 2019, irrespective of the fertility scenario. A lower (higher) than baseline fertility rate would, however, increase (decrease) the relative expenditure on health and retirement among all public spending. Despite the government Pension Fund Global financing a substantial portion of public expenditure, Norway will in all likelihood lose its previous room for fiscal manoeuvre as oil revenues are projected to decrease in the future while the pension fund is forecasted to grow slower than the mainland economy. Norwegian projections suggest that different fertility rates would not change this noticeably. To respond to future economic, fiscal pressures, Norway could therefore shift the focus away from raising fertility rates and instead ease future pressure on the pension systems by prolonging working lives, boosting economic productivity and actively encouraging increased long-term investment in private pensions.
Stylised projections suggest that an ageing Norwegian society will require that the long-term care (LTC) workforce is 110% larger than what is projected for 2060. Even when factoring in productivity increases in the LTC sector, the required increase would be around at 69%. Higher fertility rates would somewhat reduce these pressures as more individuals would flow into the labour force over time – and enter LTC professions – but this would not be sufficient to avert a looming LTC crisis in the future. The rising demand for long-term care could be met by steering a growing share of the workforce into LTC profession – particularly boys – while critically improving productivity in the LTC sector through a better use of the available health and care workforce and an increased use of technology and digital solutions. Targeted migration programmes could further support the mitigation of future shortfalls in the LTC workforce.
The following sections of this chapter look to the future and aim to project how different scenarios in the dynamics of the future fertility rate in Norway and selected OECD countries may affect economic and social outcomes by the year of 2060. This chapter considers a baseline scenario, which represents the expected development of the total fertility rate (TFR) along with a high and a low fertility scenario that illustrate the links between fertility and economic and social outcomes in a stylised way. All baseline and scenario projections are presented for Norway as well as other selected OECD members, based on relevance and data availability. This includes Norway’s Nordic neighbours Denmark, Finland, Iceland and Sweden, as well as Germany – which faces particularly high pressures related to population ageing – and France – which has been able to keep their fertility rates comparatively high. Projections for the United States are presented as well to add a non-European context to the exercise.
The demographic projections under different fertility scenarios itself require projections not only on fertility rates by age group, but also on the underlying sex- and age-disaggregated projections of mortality rates and net migration to fully model future demographic dynamics. Such projections are, for example, readily available in the Eurostat 2019 Population Projections (Eurostat, 2020[11]) and the United States 2017 National Population Projections (U.S. Census Bureau, 2017[12]). Other long-term population projections, such as the UN 2022 World Prospects (UN DESA, 2022[13]) or the Norway’s 2022 National Population Projections (Thomas and Tømmerås, 2022[14]), do not provide age- and sex-disaggregated net migration projections. Therefore, all projections in this chapter take the Eurostat 2019 Population Projections and the United States 2017 National Population Projections as the baseline for population development until 2060 and adjust estimates following a range of scenarios about future fertility dynamics (Box 6.1). As population projections are only illustrative and heavily dependent on the underlying assumptions, all outcomes must be treated with caution. They also cannot account for how current and future generations will react to the economic and social insecurity generated by the long-term consequences of the COVID-19 pandemic, the cost-of-living crisis, Russia’s war of aggression on Ukraine and the threat of irreversible climate change.
The projections on economic outcomes presented in this chapter follow the economic and fiscal frameworks of the OECD Long-Term Growth Model as well as the OECD in-house labour force participation projections (Guillemette and Turner, 2018[4]; 2021[15]; Cavalleri and Guillemette, 2017[16]). This allows to project economic outcomes based on different scenarios on the fertility rate in selected countries, including potential economic output and employment. Among other channels, the fiscal framework of the long-term model simulates primary expenditure, including public health, long-term care and pension spending, relative to GDP. This allows to project government spending under varying fertility trajectories. A detailed methodology that underlies the projections is available in Annex 6.A.
Despite their illustrative usefulness, such projections have clear limitations. For example, the OECD’s Long-Term Growth Model outputs economic production as potential GDP per capita, which refers to economic output at full employment of all members of the labour force. In addition, the model is unable to project public revenue for Norway, as the influence of the Norwegian Government Pension Fund Global (GPFG) – sometimes also referred to as Norway’s Sovereign Wealth Fund – cannot be accurately reflected in the OECD Long-Term Growth Model. In terms of public expenditure, any non-health and non-pension related expenditure has no allocation made for different levels of spending by age, so changes in the expenditure on family benefits and services for children may be underestimated. As such the estimated effects of fertility rates on public spending may not be modelled fully accurately. A more detailed discussion on the limitations can be found in Annex 6.A.
While the population projections in all scenarios are built on the Eurostat baseline scenario or the US Population Projections main series, the specific alternative scenarios considered in this exercise integrate the high- and low fertility scenario following a similar methodology to the one used on the UN 2022 World Population Prospects (UN DESA, 2022[13]), while keeping the other baseline dynamics of the Eurostat and US projections (e.g. migration and deaths by age groups). The reason for this is the lack of sufficient age-specific fertility rates under both high and low scenarios in the Eurostat and US population projections. In each case, the divergence from the Eurostat and US fertility rates starts in 2023. It then projects TFRs that either converge to 0.5 above or below the TFR in the baseline scenario. Over the initial years of the projections, TFRs are assumed to converge slowly up- or downward to this level, following similar convergence as in the UN 2022 World Population Prospects. The precise scenarios are the following:
Baseline: age-specific fertility rates follow the baseline scenario of the Eurostat 2019 and US 2017 Population Projections. Eurostat TFR projections are based on a continuation of recent fertility trends as well as a long-term convergence towards a TFR of 1.83 by 2 100 (i.e. the UN’s World Population Prospects 2019 maximum TFR for 2 100 among all countries included in the Eurostat projections). This projects an assumed slow rebounding of fertility through a long-term catching up of postponed births (Eurostat, 2020[11]). US Census Bureau TFR projections assume that by 2 100, the age-specific TFRs of all ethnic and racial population groups in the country converge to the average age-specific fertility rates of the US-born Caucasian population for the years 2004 to 2015 (U.S. Census Bureau, 2017[12]).
High fertility: for each country, age-specific fertility rates under the baseline scenario are adjusted upward so that the TFR increases by 0.25 between 2024 and 2026, by 0.40 between 2027 and 2031, and by 0.5 from 2032 to 2060.
Low fertility: for each country, age-specific fertility rates under the baseline scenario are adjusted downward so that the TFR decreases by 0.25 between 2024 and 2026, by 0.40 between 2027 and 2031, and by 0.5 from 2032 to 2060.
Such alternative scenarios are by no means likely future fertility trajectories, and they may even represent more extreme outcomes. At the same time, for some countries – notably Finland, Iceland and Norway – declines in the historical fertility rates between 2009 and 2020 come reasonably close to the 0.5 drop in the low fertility scenario. Especially the low fertility scenario may represent an additional fertility decline similar to previous trends. In any case, the assumed clear divergence from baseline trajectories of fertility rates in the alternative scenarios serve as illustrative ‘what-if projections’ that showcase the impact markedly lower- or higher than baseline fertility could have on demographic-, economic- and social outcome measures.
Figure 6.1 plots the dynamics in the total fertility rate (TFR) under the three different scenarios, along with historical TFRs since 1960 for perspective (see Annex Figure 6.B.1 for countries other than Norway). For Norway, Eurostat projects marginal increases in the TFR from 1.52 in 2020 to 1.62 in 2060, the level previously attained in 2017. Under the low fertility scenario, the Norwegian TFR would fall initially to 1.06 by 2032, which is comparable to the fall in fertility observed between 2009 and 2020 or 1965 and 1977. After this initial decrease, the TFR in the low fertility scenario would follow a parallel trend to the slowly increasing baseline TFR, reaching a level of 1.12 in 2060. The high fertility scenario would see an initially steep increase in the TFR to 2.06 in 2035, the level last attained in 1974, before reaching a TFR 2.12 in 2060. The dynamics of the TFR scenarios for the other considered countries have similar divergence from the baseline fertility projections, though some falls of the TFRs under a low scenario are reaching a particularly low level (e.g. temporarily below a TFR of 1 in Finland). As all the scenarios, including the baseline, are highly stylised, there is a lot less volatility in the projections than what was visible in the past. Indeed, past fertility rates may have continuously reacted to changes in business cycles, global events or other factors, which cannot be accounted for in future projections.
Note: The baseline scenario refers to baseline fertility rates in the Eurostat population projections, which converge to a TFR of 1.62 by 2060 for Norway. “Low fertility” refers to fertility rates that converge to a TFR 0.5 below baseline. “High fertility” refers to fertility rates that converge to a TFR 0.5 above baseline.
Source: OECD Family Database and OECD calculations based on Eurostat Population Projections 2019.
Demographic projections are built upon a range of assumptions over factors that directly influence the size of the population over time. This generally includes the future dynamics of fertility and mortality rates as well as the in- and outflow of individuals through migration, all disaggregated by age groups and sex. As such, all outcomes of such projection exercises are highly dependent on assumptions and therefore must be seen as illustrative only and interpreted with caution.
The population projections this chapter is built on are mainly based on demographic forecasts of Eurostat, enriched with fertility scenarios from the United Nations, rather than population projections prepared by national statistical institutes (NSIs), such as Statistics Norway. The projections for each presented country follow a common methodology and allow to compare the demographic development of countries. For the Eurostat projections in particular, convergence of socio-economic differentials between countries is one of the central assumptions, and age-specific fertility, mortality and migration rates are converging over the long-term. In the case of fertility, for example, this means that the total fertility rates (TFRs) are converging to the maximum TFR among EEA countries that UN’s World Population Prospects 2019 project in 2 100, namely a TFR of 1.83. Any differences in fertility rates between the countries are modelled to be fading out in the very long term.
Source: Eurostat Population Projections 2019, Statistics Norway National Population Projections 2020 and Statistics Norway National Population Projections 2022.
Based on different approaches and assumptions that Eurostat and other institutes may take, some of these projections may differ from each other. For example, for Norway, the baseline population projections of Eurostat assume a notably higher net migration and mortality rate than Statistics Norway does in their own projections from 2020 and 2022. In addition, Statistics Norway assumes a much faster and higher rebounding of fertility (Figure 6.2).
Prior to the 2020 population projection, Statistics Norway often substantially underestimated net migration rates, while moderately overestimating the fertility development. This may have spurred upward adjustments of migration rates in recent projections, which thus exceed projections of Eurostat. The accuracy of the 2020 population projection for 2020 and 2021 is hard to gauge due to the effect of the COVID-19 pandemic on all of the demographic inputs. With 6.1 million, however, Statistics Norway projects a population for 2060 that is noticeably smaller than what is projected by Eurostat (6.54 million).
Source: Eurostat (2020[11]), United Nations (2019[17]), Syse, Thomas and Gleditsch (2020[18]), Gleditsch, Rogne and Thomas (2021[19]), Thomas and Tømmerås (2022[14]).
The potential impact fertility rates on economic and social outcomes are manifold, but the most direct effects are present in demographic outcomes. Intuitively, a low fertility scenario, for example, would decrease the number of births relative to the baseline projections. This would have several direct implications from the year after fertility falls below baseline and change the demographic structure of a country over time. For Norway, even the baseline fertility scenario projects notable demographic changes as the population gradually ages (Figure 6.3). Compared to 2020, Norway’s population in 2060 will consist of a substantially larger share of elderly (aged 65 or above), increasing from 18% to about 26% of the total population. At the same time, the population of the young (below the age of 15) will decrease slightly from 17% to 14%, while the working age population (aged 15-64) is projected to decrease from 65% to 59% by 2060.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. Other countries in Annex 6.B.
Source: OECD calculations based on Eurostat Population Projections 2019.
The demographic structure in 2060 is sensitive to different scenarios on fertility rates. Under the low fertility scenario, the share of the young would thin out stronger than under the baseline projections, falling to 11% of the entire Norwegian population with a trend that would likely exacerbate this development further beyond 2060. At the same time, the elderly population would represent 29% of the Norwegian population, while the working age population would be marginally larger than under the baseline projection in 2060 (60%). General ageing of the Norwegian population could be halted under the high fertility scenario, which would slightly increase the population of the young to 18% and limit the increase of the elderly population by 2060 to 24%. With 59%, the working age population would make up a similar part of the total Norwegian population as under the baseline scenario.
Most of the other considered countries would see very similar dynamics to Norway, though mostly with slightly slower population ageing (Annex Figure 6.B.2). However, Iceland stands out with a particularly strong increase in the elderly population at baseline between 2020 and 2060 (+10%), as it is a relatively “young” population at present. Even under the high fertility scenario it would not be able to avert these trends as the share of the elderly population would still increase by 8%, while the share of the young would remain roughly stable. Should Iceland face a low fertility path, it may even expect an increase of 13% in the share of the elderly population. Nevertheless, with one of the lowest median ages across the OECD, Iceland may still be better equipped to handle such pressures of population ageing than many other countries (OECD, 2019[20]). Indeed, even with the strongest increase in the share of the elderly population, it would still remain the youngest country of those considered in this report.
A growing share of the elderly population, with a simultaneous decrease in the share of the working age population, as observed for all considered countries and across all scenarios, will mean that in relative terms, fewer working individuals will have to support more and more people beyond retirement age. Norway will also experience this future, but likely to a lesser extent. The old-age dependency ratio (i.e. the ratio between the elderly and working age populations) shows that in 2020, there were about 27 elderly individuals per 100 persons of working-wage in Norway (Figure 6.4). Between 2020 and 2060, Norway’s old-age dependency ratio is set to increase to 44% at baseline, which is the second largest relative increase projected across the considered countries. Going from 22% to 41% between 2020 and 2060, only Iceland is expected to see a bigger increase. However, both Iceland and Norway still have relatively low old-age dependency ratios at the onset of the 2020s, and therefore won’t reach the same levels as in Finland, France and Germany, where approximately 50 elderly individuals per 100 persons of working-age or more are projected by 2060.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. The old-age to working-age demographic ratio is defined as the number of individuals aged 65 and over per 100 people of working age defined as those at ages 20 to 64.
Source: OECD calculations based on Eurostat Population Projections 2019 and United States 2017 National Population Projections.
As expected, based on the demographic development under the different scenarios discussed above, lower than baseline fertility rates would further increase demographic pressures by 2060. However, Norway would only have to expect a small difference in the ratio of elderly per 100 persons of working age under the baseline (44 persons) and low fertility scenarios (48 persons). At the same time, the country could benefit notable from high fertility rates, which could push the old-age dependency ratio just under 40 senior citizens per 100 individuals of working age. All else equal, all these scenarios will not only result in a growing public spending on retirement incomes, but overall health expenditures may substantially rise as well (OECD, 2016[21]). The fiscal pressure caused by the general population ageing will be discussed in Section 5.5 below.
All else equal, fertility induced changes in population dynamics and the demographic structure of a country will also affect the inflow of people into the labour force. For example, should a lower fertility scenario lead to fewer births, then the number of individuals entering the labour force would be reduced relative to baseline once children come of age. This means that from 15 years after the scenarios diverge, the size of the labour force would diverge under different assumptions on fertility as well. Indeed, when considering the development of the Norwegian labour force, it remains stable up until the late 2030s, while it starts to diverge under the two alternative fertility scenarios afterwards (Figure 6.5). Like the overall dynamics of the working-age population, the high fertility scenario would have a large effect on the aggregate labour force, growing 18% by 2060 relative to 2020, compared to 10% under the baseline projection. However, even the low fertility scenario still projects a slight 1% increase of the labour force relative to 2020 (Figure 6.6).
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. Other countries in Annex 6.B.
Source: OECD calculations based on OECD in-house labour force projections, Eurostat Historical Population Projections 2019 and United States 2017 National Population Projections.
Similar to Norway, Iceland and Sweden are also expected to see continuous growth in their headcount labour force, while those in Denmark, Finland, France and Germany are set to decrease by 2060 (Annex Figure 6.B.3). However, even under a low fertility scenario, Iceland, which is by far the youngest country considered in this chapter, is projected to see higher growth in the labour force than in Norway under the high fertility scenarios (Figure 6.6). At the same time, Denmark and France would be able to avert their projected decline in the labour force with high fertility, though this would only reach 2020 levels by the mid-2050s. Thus, any positive divergence from the baseline fertility path would only pay off well into the future. However, the positive growth dividend beyond 2060 might potentially be large.
As the only factors contributing to potential economic output in the OECD Long-Term Growth model are capital, productivity, and potential employment (see more in Annex 6.B) – and while the latter is the only one that is directly affected by differences in fertility rates – it is only the size of the labour force that will induce changes in the total production under the different fertility scenarios. Aggregate production, i.e. the potential gross domestic product (potential GDP), would be stable for at least 15 years while the labour force remains unchanged before fertility would have any effects. However, given changes in the size of the population that are effective right from the start of the divergence in fertility rates, economic output would be shared by a different population base. For example, in a lower fertility scenario, potential GDP would initially stay stable, while the size of the population starts to negatively diverge from the baseline projection. This would initially lead to increased GDP per capita. However, over time fewer would enter the labour force, which would slowly reduce aggregate output and eventually negatively impact GDP per capita as well.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios.
Source: OECD calculations based on OECD in-house labour force projections, Eurostat Historical Population Projections 2019 and United States 2017 National Population Projections.
Indeed, all countries that are considered in this chapter exhibit an initial positive effect on potential GDP per capita growth under the low fertility scenario relative to the baseline projection, as economic output is shared among fewer people overall (Figure 6.7). Reaching the late 2040s however, all countries cross over from a positive effect on potential GDP per capita growth under low fertility, to a lower annual growth in GDP per capita than under the baseline projection. In some countries, the difference between annual per capita growth in potential GDP per capita under low and baseline fertility is rather small beyond the cross-over points, such as in Denmark, Iceland, Norway, Sweden, and the United States. For other countries, specifically France and Finland, this difference is somewhat larger, slowly eroding the positive annual per capita growth effects of low fertility over time. It is complicated to pinpoint the exact reasons for different effects for different countries, as the per capita growth projections are formed by a variety of input factors – such as the dynamics in demographic structures, labour force participation rates for men and women as well as across different ages, and productivity projections (see more in Annex 6.A).
Note: See note to Figure 6.2. for a definition of the different fertility scenarios.
Source: OECD calculations based on Eurostat Historical Population Projections 2019, United States 2017 National Population Projections and the OECD Long-Term Growth Model.
In most of the countries, the initial boost to annual per capita growth in potential GDP was comparatively large however, so it the aggregate effect of low fertility on per capita growth until 2060 is still positive for all (Figure 6.8). Above the average annual growth rate of 0.81% in potential GDP per capita under the baseline projection, Norway would, for example, see an average of additional 0.09 percentage points of annual potential GDP per capita growth under the low fertility scenario – an effect of the same size that could be expected when fully closing the gender gap in labour force participation by 2060 (see OECD (2022[22])). The largest effects, an average of additional 0.10 or more percentage points of annual growth in potential GDP per capita under low fertility could be expected in Sweden and the United States (Figure 6.8). Higher fertility, on the other hand, would reduce average annual GDP per capita growth by 2060, for most countries even stronger than low fertility increased growth. In the case of Norway, for example, average annual growth would shrink by 0.08 percentage points.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios.
Source: OECD calculations based on Eurostat Historical Population Projections 2019, United States 2017 National Population Projections and the OECD Long-Term Growth Model.
The nature of the fiscal outcomes is highly dependent of the demographic structure of a country. For example, an increasing share of the elderly among the overall population will result in a growing expenditure on retirement income, while overall health expenditures may rise as well as they substantially increase with age (OECD, 2016[21]). If the working-age population shrinks at the same time, while other factors remain the same, public revenue would decrease as the population paying income tax shrinks with the size of the labour force. As such, ageing populations will in all likelihood face an increased fiscal burden over the coming decades, for example requiring adjustments in the national pensions system and the Tax-Benefit system as a whole to remain fiscally sustainable (Guillemette and Turner, 2021[15]).
6.5.1. Government expenditure
Given the importance of fertility rates for population ageing (see above), they are an important factor for public primary expenditure – government spending excluding any interest payments – both in the short- and long-term. While not immediately affecting expenditure on pension and health, changes in fertility rates will have particular impact on other primary government expenditure, such as spending on early childhood education and care (ECEC), the school system as well as family benefits and allowances. A lower fertility rate should initially induce lower public expenditure as fewer children are born and eventually enter the school system, impacting spending both in absolute and relative terms while the elderly population remains stable.
With the long time between birth and pension entry, any change in fertility rates would not have a direct effect on absolute pension spending over a projection horizon ending in 2060. However, despite the absence of immediate direct effects, there are indirect effects on the relative share of total pension expenditure. As such, decreases in the young population would mean that, in relative terms, the share of pension spending among the total public expenditure increases. Once a lower fertility rate would lead to decreases in the labour force relative to baseline, aggregate economic output would also fall, further increasing pension expenditure expressed as a share of GDP. At the same time, an older population would have to direct more of the total public expenditure to health and long-term care in relative terms.
Indeed, as shown in Figure 6.9 for Norway and Annex Figure 6.B.4 for other countries, government primary expenditure is set to increase irrespective of the fertility scenarios (except for the United States, which nevertheless has continuous expenditure growth beyond 2022). However, lower fertility rates would initially decrease overall primary government expenditure relative to the baseline, as the size of the population shrinks, while higher fertility rates would increase public spending. The magnitude of this effect is relatively similar across countries for all fertility scenarios. As the effect of other primary revenue is underestimated, however, it is almost certain that the downward pressure on primary revenue through fertility would be stronger than projected, as expenditure in ECEC and the school system is substantial. Like the underestimation of the negative effects of low fertility on primary expenditure, the true increase of spending under high fertility will likely be larger.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. Primary expenditure is defined as government spending on pensions, health, long-term care and other categories(e.g. including family transfers and in-kind services), excluding any expenditure on interest payments. Other countries in Annex 6.B.
Source: OECD calculations based on Eurostat Historical Population Projections 2019 and the OECD Long-Term Growth Model.
Much of these effects of higher and lower fertility rates would slowly reverse from the early 2040 onwards, converging close to baseline level expenditure by the late 2050s. However, while the difference in total public expenditure in 2060 between all three scenarios in Denmark, Iceland and Sweden is marginal at best, lower fertility would increase public expenditure from this point onwards in Finland, France, Germany, Norway, and the United States. Even though the projection horizon does not allow to assess the dynamics under the different scenarios beyond 2060, the presented trends over the late 2050s could exacerbate expenditure increases driven by low fertility beyond 2060.
Irrespective of the convergence of total primary public expenditure under different fertility scenarios for Norway, the drivers of these changes are not the same under the different projections (Figure 6.10). Most changes in total primary expenditure between 2019 and 2060, which is projected to rise by 6.0 percentage points relative to GDP under the baseline scenarios, can be attributed to changes in health spending (+3.0 percentage points). However, the low fertility scenario, which would only see slightly larger overall expenditure (+6.4 percentage points), would lead to a somewhat stronger increase in health spending (+3.4 percentage points) as well as a notable increase in pension spending (+2.4 percentage points), while other primary expenditure would marginally increase (+0.5 percentage points). Higher fertility rates would only slightly dampen the increase in overall expenditure to 5.8 percentage points.
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. Primary expenditure is defined as government spending on pensions, health, long-term care and other categories (e.g. including family transfers and in-kind services), excluding any expenditure on interest payments. The figure compares primary public expenditure in 2060 to 2019, in contrast to 2020 as most other figures. The reason for this is the substantial expenditure on fiscal relief packages over the course of the COVID-19 pandemic, which increased public expenditure substantially, most notably for the United States. While other countries, for example and Denmark and Germany, provided large guarantees and loans, such items do not add to primary expenditure, unlike direct additional spending as in the United States (IMF Fiscal Affairs Department, 2021[23]). Note that the “Other” category of public expenditure assumes that governments will seek to provide a constant level of non-health/non-pension per capita spending in real terms over time. No allocation is made here for different levels of spending by age. As a result, the change in this category is likely underestimated. See more in Guillemette and Turner (2021[15]; 2017[24]) and Annex 6.A.
Source: OECD calculations based on Eurostat Historical Population Projections 2019, United States 2017 National Population Projections and the OECD Long-Term Growth Model.
Other countries see generally comparable changes in the different sub-components of primary public expenditure (Figure 6.10). However, for Denmark and Sweden, the projections under different fertility rates do not lead to noticeable changes in overall expenditure. Here, the lower increases in pension and health expenditure under higher fertility rates, for example, exactly offset the higher increases in other primary expenditure (e.g. family benefits and expenditure on ECEC and education). In Finland and France, increases are particularly strong on pension expenditures under low fertility, raising by more than 5.0 percentage points relative to GDP. Iceland is the only country in which lower fertility would lead to noticeably lower increases in public expenditure (-0.2 percentage points relative to baseline), even though it has by far the strongest increase at baseline (8.7 percentage points). Much of this comes through increases in heath expenditure, which raises by 3.6-4.3 percentage points across the different scenarios.
6.5.2. Government revenue
There is good reason to assume that primary government revenue is on a clear and long-term downward trend. For several years, growth in tax revenues has been noticeably decreasing, in part as a result of expensive incentive schemes to steer the population towards more environmental sustainability, such as wide-scale tax and toll exemptions for the purchase of zero-emission and hybrid cars, in contrast to the heavy taxation of those running solely on fuel (see e.g. Eskeland and Yan (2021[25])). In 2022, revenue from car related taxation is estimated to be about half of what it was just 15 years ago, falling by about NOK 40 billion (USD 4.04 billion) or about 6% of total tax revenue. Even though some vehicle taxes and tolls are set to being re-introduced, particularly for hybrid cars, this is unlikely to fill the fiscal gap created by wide-spread adoption of zero- and low-emission cars. Slowing productivity and labour force growth have also contributed to the expectation of a growing fiscal gap (Norwegian Ministry of Finance, 2021[5]; OECD, 2022[26]).
At the same time, Norway has been heavily reliant on the Government Pension Fund Global (GPFG) as an important element of the fiscal framework (Box 6.2). Since its creation in the early 1990s, it has aimed for an intergenerationally fair use of petroleum revenues and potentially allows to offset some of the negative fiscal effects of population ageing by curtailing the need for increased tax revenue or spending cuts compared other countries considered in this chapter (OECD, 2021[27]). Norway is therefore in a fortunate position where it can run substantial “non-oil” fiscal deficits that are cushioned by the regular withdrawal of the expected annual returns of the GPFG. However, decreasing oil revenues could substantially reduce the fiscal space created by the GPFG in the future and thus require more efficient public spending as Norway’s ratio of public expenditure to GDP is also among the highest in the OECD (OECD, 2022[26]). Importantly, revenues from the GPFG are not included in the Norwegian Government’s primary revenue as they purely consist of interest and investment income. As such, they do not enter primary balance calculations, even though they are generally used to cover the structural “non-oil” deficit.
With projected declines in the GPFG as well as it susceptibility to global macroeconomic risks, the relative contribution of the GFPG to financing government spending will likely decline in the future. Indeed, the overall value of the GFPG is projected to grow more slowly in the future than over past decades as petroleum revenues are on a declining trend (Norwegian Ministry of Finance, 2021[5]; OECD, 2022[26]). The Norwegian non-oil economy will thus eventually grow faster than the fund itself. In terms of the funds value relative to Norwegian GDP, it is initially projected to increase somewhat, reaching a height of almost 380% of GDP in 2030, but it is projected to be about 270% of GDP in 2060. While this alone will reduce the fiscal space offered by the GPFG, the value of the fund is also highly sensitive to unpredictable macroeconomic developments (Box 6.2 and Figure 6.11).
The Norwegian Ministry of Finance expects that, without substantial reform, the average annual growth in tax revenue is set to fall from NOK 18 billion (USD 1.82 billion) between 2011 and 2019 to NOK 10 billion (USD 1.01 billion) between 2023 and 2030 (Figure 6.12). The Norwegian Ministry of Finance projects that the overall returns from taxation and GPFG withdrawal will be just enough to cover the expenses on the national insurance and pension system, eradicating the fiscal space for other prioritised policy initiatives that have been common over recent decades (Norwegian Ministry of Finance, 2021[5]).
With revenues from offshore petroleum production, Norway has accumulated substantial financial assets in the Government Pension Fund Global (GPFG) – sometimes also referred to as Norway’s Sovereign Wealth Fund – based on the net cash flow and financial transactions from the petroleum sector as well as the regular returns on the fund’s assets. The fund, managed by Norges Bank Investment Management, has a value equivalent to around 3.5 times annual GDP. Foreign assets make up the entirety of its investments – about 70% equities with the remainder in bonds and property – which help offset the currency appreciation arising from petroleum exports.
Under the Norwegian fiscal framework, withdrawal of the expected real return from the fund, currently set at 3% of the funds volume, covers the cyclically adjusted non-oil deficit (the “structural non-oil deficit”), at-present in the order of 10% of GDP. The framework implies an intergenerationally fair use of oil wealth because withdrawal of expected returns does not deplete the fund over time. Nevertheless, business cycle considerations are given significant emphasis which can lead the actual takeout rate to deviate from the expected real return both from one year to the next and over several years.
Note: See OECD (2022[26]) for detailed information.
Source: OECD (2022[26]), based on calculations drawing on data from the Norwegian Ministry of Finance.
The GPFG is highly sensitive to global macroeconomic developments, such as a changing oil and gas revenues as well as the volatility in the returns on its assets. Even though the surge in oil and gas prices following the Russia’s war of aggression on Ukraine will increase the revenue flow to the fund temporarily, this is unlikely to be a permanent effect as many countries scramble to scale up alternatives to their current reliance on oil and gas in industry and private consumption.
Source: OECD (2022[26]), Norwegian Ministry of Finance (2021[5]), Norges Bank (2022[28]).
However, a continuation of recent downward trends in fertility could slightly ease the pressure on the Norwegian primary balance, at least over a mid-term horizon. Projections by the Norwegian Ministry of Finance (2021[5]) and Statistics Norway (2019[6]) show that lower than baseline fertility rates, for example, would decrease the fiscal gap in the future (these projections use Statistics Norway’s own population and fertility projections). Lower than baseline fertility would initially reduce the primary deficit by more than 6 percent in the early 2040s, before increasing it by about 2 percent towards the end of the 21st century. Overall, fertility seems to be less important for the Norwegian fiscal framework than external macroeconomic factors.
Note: “Demography” is an estimate of the increasing healthcare costs due to population ageing. “National Insurance Scheme” mainly reflects increasing costs in pensions and disability benefits. The NOK values in the calculations are re-based to 2021 and therefore 2021 mainland GDP is the denominator.
Source: OECD (2022[26]) based on Ministry of Finance and OECD calculations.
As such, fiscal pressure may mount, generally independent of future fertility rates. Instead of aiming for higher fertility rates to avert economic and social pressures, prolonging working lives, increasing long-term investment in private pension savings and improving productivity might thus be more effective. Fewer births than in previous decades are therefore not necessarily a serious fiscal and economic concern, as long as Norway prepares for the future and keeps its population well informed about necessary adjustment in previous policies.
In addition to rising expenditure on health services, an ageing society is likely to experience an increased demand for long-term care (LTC) services for the elderly. It is likely that general trends of increasing life expectancy will coincide with increases in healthy life expectancy (Foreman et al., 2018[29]), but at the same time, more and more senior citizens suffer from multiple chronic conditions that require specialised and intensive care, while comparatively low-pay and ever more stressful jobs limit retention in the LTC workforce (OECD, 2020[30]). The question of how the demand and supply of LTC services will evolve in the future is thus impossible to answer with any degree of certainty.
This section analyses the demand and supply of long-term care services in Norway, based on the demographic developments under the three fertility scenarios introduced earlier in this chapter. While fertility trajectories do not have any direct effects on the number of senior citizens that may need long-term care services in the distant future, fertility will critically shape the size of the labour force in the decades to come (see above) and may thus be a factor in ensuring adequate care provision for future generations of elderly. By keeping the ratio between care workers and LTC users as well as the share of LTC workers among the labour force constant – and also accounting for increases in (healthy) life expectancy – these projections provide simple illustrative scenario on the gap between LTC demand and supply by 2060 (a more comprehensive approach on modelling future LTC demand and supply is taken in OECD (Forthcoming[31])). A detailed methodology is available in Annex 6.A.
With an increasing population of elderly, and in the absence of technological advances, Norway would need noticeably more LTC workers in the future, a trend that would worsen continuously over the projection horizon. In 2060, Norway would need approximately 266 400 LTC workers that would care for 391 700 senior citizens with support needs. While this would require an increase in the LTC workforce of about 135%, the baseline fertility scenario projects a mere 10% increase to 126 000 LTC workers by 2060. The substantial growth in the elderly population and the modest increase in the Norwegian labour force would therefore lead to a substantial mismatch between the size of the projected LTC workforce and how many LTC workers are needed to provide a continuum of today’s quality of care – with the required LTC workforce being more than twice as large as the one that is projected (Figure 6.13).
Note: See note to Figure 6.2. for a definition of the different fertility scenarios. The calculations assume that the ratio between LTC workers and elderly LTC users remains unchanged between 2019 and 2060, while the same share of the overall workforce is employed in the LTC sector in both years. Productivity increases are assumed to reduce the ratio between LTC workers and elderly in 2060, based on similar approach as in OECD (2020[30]). See more in Annex 6.A.
Source: OECD calculations based on Eurostat Historical Population Projections 2019, Statistics Norway (2022[32]) and OECD (2021[33]; 2020[30]).
While changes in fertility won’t influence the demand for long-term care over the projection horizon, it will directly influence the supply of LTC workers through its effects on the size of the labour force. Lower than baseline fertility, for example, will not only decrease the ratio of the working age population to the elderly, but it will also reduce the number of LTC workers. Higher fertility rates may be able to dampen the effects of a looming LTC crisis in the future, but as their effects on the size of the labour force only materialise with a delay, they would only change the LTC care gap from around 2045 onward. However, even under higher fertility rates, such LTC care gaps would not substantially diminish, only dropping to a relative gap of 91% in 2060 (Figure 6.13).
However, it may be reasonable to expect some sort of productivity increase in the LTC sector over the coming decades – for example through assistive robots and other technologies – which would mean that each worker could care for more people without compromising quality. The Norwegian Helsepersonellkommisjonen (Health Personnel Commission) also emphasises the increased use of technology and digital solutions (see also Savage (2022[34])) and the need for a better use of the available health and care workforce and, rather than aiming for an increase of the workforce, to avert a looming LTC crisis (Ministry of Health and Welfare, 2023[10]). In 2019, Norway had one of the highest ratios between the elderly and the LTC workforce (see OECD (2021[33])), thus there may be scope for reduction in this ratio.
The projections Figure 6.13 provide an additional scenario that factor in productivity increases, following OECD (2020[30]) (see more in Annex 6.A). Even when assuming a reasonable boost to the productivity of the health and care sector – which typically exhibits slower productivity growth than the aggregate economy – the gap in the LTC workforce would remain large. In this case, Norway would still require 69% more LTC workers than projected by 2060. Should productivity increases alone avert a looming LTC crisis (see Ministry of Health and Welfare (2023[10])), it would require a more substantial reduction in the ratio between LTC workers and users. For example, in the stylised model developed here, this would require a reduction from 68 to about 32 LTC workers per 100 LTC users at baseline by 2060 – a necessary productivity increase of more than 50%, and thus substantially stronger than any productivity boost expected for the aggregate Norwegian economy. Even under the higher fertility scenario, Norway would have to increase productivity in the LTC sector by 47% (to 36 LTC workers per 100 LTC users in 2060). Such efforts would likely also result in a reduced quality of elderly care and thus may be highly undesirable for the country.
Based on the stylised results of the projection exercise, boosting fertility rates and productivity increases alone does not seem to be fully effective in reducing future LTC gaps. While productivity increases in the LTC and health sector are necessary, they alone are unlikely to avert a substantial gap in the demand and supply of LTC. In addition to promoting healthier lives, a better utilisation of the available health and care workforce and general productivity increases, Norway could aim for an expansion of the LTC workforce by steering more into training and education related to geriatric care – including boys and men, who hold less than 10% of jobs in the LTC sector in Norway (OECD, 2020[30]) – or specific migration channels to recruit foreign LTC workers, in order to avert a looming LTC crisis in the future.
References
[16] Cavalleri, M. and Y. Guillemette (2017), “A revised approach to trend employment projections in long-term scenarios”, OECD Economics Department Working Papers, No. 1384, OECD Publishing, Paris, https://doi.org/10.1787/075f0153-en.
[37] Chłon-Dominczak, A. et al. (2019), “Welfare state and the age distribution of public consumption and public transfers in the EU countries”, Vienna Yearbook of Population Research, Vol. 1, pp. 071-097, https://doi.org/10.1553/populationyearbook2019s071.
[3] Crowe, D. et al. (2022), “Population ageing and government revenue: Expected trends and policy considerations to boost revenue”, OECD Economics Department Working Papers, No. 1737, OECD Publishing, Paris, https://doi.org/10.1787/9ce9e8e3-en.
[25] Eskeland, G. and S. Yan (2021), “The Norwegian CO2-differentiated motor vehicle registration tax: An extended Cost-Benefit Analysis”, OECD Environment Working Papers, No. 178, OECD Publishing, Paris, https://doi.org/10.1787/ee108c96-en.
[38] European Commission (2021), The 2021 Ageing Report Economic & Budgetary Projections for the EU Member States (2019-2070), https://ec.europa.eu/info/sites/default/files/economy-finance/ip148_en.pdf.
[11] Eurostat (2020), Methodology of the Eurostat population projections 2019-based, https://ec.europa.eu/eurostat/cache/metadata/Annexes/proj_esms_an1.pdf.
[29] Foreman, K. et al. (2018), “Forecasting life expectancy, years of life lost, and all-cause and cause-specific mortality for 250 causes of death: reference and alternative scenarios for 2016–40 for 195 countries and territories”, The Lancet, Vol. 392/10159, pp. 2052-2090, https://doi.org/10.1016/s0140-6736(18)31694-5.
[9] Gietel-Basten, S., A. Rotkirch and T. Sobotka (2022), “Changing the perspective on low birth rates: why simplistic solutions won’t work”, BMJ, p. e072670, https://doi.org/10.1136/bmj-2022-072670.
[19] Gleditsch, R., A. Rogne and M. Thomas (2021), The accuracy of Statistics Norway’s national population projections, https://www.ssb.no/en/forskning/discussion-papers/_attachment/449813?_ts=17868b4f140.
[7] Goldin, I. (2022), “Demography is not destiny”, Financial Times, https://www.ft.com/content/e04ba005-a913-4362-8434-dae488220310.
[36] Guillemette, Y. (2019), “Recent improvements to the public finance block of the OECD’s long-term global model”, OECD Economics Department Working Papers, No. 1581, OECD Publishing, Paris, https://doi.org/10.1787/4f07fb8d-en.
[35] Guillemette, Y., A. De Mauro and D. Turner (2018), “Saving, investment, capital stock and current account projections in long-term scenarios”, OECD Economics Department Working Papers, No. 1461, OECD Publishing, Paris, https://doi.org/10.1787/aa519fc9-en.
[39] Guillemette, Y. et al. (2017), “A revised approach to productivity convergence in long-term scenarios”, OECD Economics Department Working Papers, No. 1385, OECD Publishing, Paris, https://doi.org/10.1787/0b8947e3-en.
[15] Guillemette, Y. and D. Turner (2021), “The long game: Fiscal outlooks to 2060 underline need for structural reform”, OECD Economic Policy Papers, No. 29, OECD Publishing, Paris, https://doi.org/10.1787/a112307e-en.
[4] Guillemette, Y. and D. Turner (2018), The Long View: Scenarios for the World Economy to 2060, https://doi.org/10.1787/2226583X.
[24] Guillemette, Y. and D. Turner (2017), “The fiscal projection framework in long-term scenarios”, OECD Economics Department Working Papers, No. 1440, OECD Publishing, Paris, https://doi.org/10.1787/8eddfa18-en.
[23] IMF Fiscal Affairs Department (2021), Fiscal Monitor Database of Country Fiscal Measures in Response to the COVID-19 Pandemic, https://www.imf.org/en/Topics/imf-and-covid19/Fiscal-Policies-Database-in-Response-to-COVID-19.
[10] Ministry of Health and Welfare (2023), Tid for handling — Personellet i en bærekraftig helse- og omsorgstjeneste, https://www.regjeringen.no/no/dep/hod/id421/.
[28] Norges Bank (2022), Market value, https://www.nbim.no/en/the-fund/Market-Value/.
[5] Norwegian Ministry of Finance (2021), “Long-term Perspectives on the Norwegian Economy 2021”, Meld. St. 14 (2020 – 2021) Report to the Storting (white paper), https://www.regjeringen.no/contentassets/91bdfca9231d45408e8107a703fee790/en-gb/pdfs/stm202020210014000engpdfs.pdf.
[26] OECD (2022), OECD Economic Surveys: Norway 2022, OECD Publishing, Paris, https://doi.org/10.1787/df7b87ab-en.
[22] OECD (2022), “Report on the Implementation of the OECD Gender”, https://www.oecd.org/mcm/Implementation-OECD-Gender-Recommendations.pdf.
[33] OECD (2021), “Long-term care workers per 100 people aged 65 and over, 2011 and 2019 (or nearest year)”, in Health at a Glance 2021: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ae3016b9-en.
[27] OECD (2021), Taxing Wages 2021, OECD Publishing, Paris, https://doi.org/10.1787/83a87978-en.
[30] OECD (2020), Who Cares? Attracting and Retaining Care Workers for the Elderly, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/92c0ef68-en.
[20] OECD (2019), Working Better with Age, Ageing and Employment Policies, OECD Publishing, Paris, https://doi.org/10.1787/c4d4f66a-en.
[1] OECD (2017), Preventing Ageing Unequally, OECD Publishing, Paris, https://doi.org/10.1787/9789264279087-en.
[21] OECD (2016), “Health Spending: Expenditure by disease, age and gender”, https://www.oecd.org/health/Expenditure-by-disease-age-and-gender-FOCUS-April2016.pdf.
[31] OECD (Forthcoming), Beyond Applause? Improving Working Conditions in Long-Term Care, OECD Publishing, Paris, https://doi.org/10.1787/27d33ab3-en.
[2] Rouzet, D. et al. (2019), “Fiscal challenges and inclusive growth in ageing societies”, OECD Economic Policy Papers, No. 27, OECD Publishing, Paris, https://doi.org/10.1787/c553d8d2-en.
[34] Savage, N. (2022), “Robots rise to meet the challenge of caring for old people”, Nature, Vol. 601/7893, pp. S8-S10, https://doi.org/10.1038/d41586-022-00072-z.
[8] Skirbekk, V. (2022), Decline and Prosper!, Springer International Publishing, Cham, https://doi.org/10.1007/978-3-030-91611-4.
[32] Statistics Norway (2022), Care services, https://www.ssb.no/en/helse/helsetjenester/statistikk/sjukeheimar-heimetenester-og-andre-omsorgstenester.
[6] Statistics Norway (2019), Langsiktige virkninger på offentlige finanser og verdiskapning av endringer i fruktbarhet, https://www.ssb.no/en/nasjonalregnskap-og-konjunkturer/artikler-og-publikasjoner/long-run-effects-on-government-finances-and-national-income-in-norway-of-changes-in-fertility.
[18] Syse, A., M. Thomas and R. Gleditsch (2020), Norway’s 2020 population projections, https://www.ssb.no/en/befolkning/artikler-og-publikasjoner/_attachment/422993?_ts=172758d6808.
[14] Thomas, M. and A. Tømmerås (2022), Norway’s 2022 national population projections, https://www.ssb.no/en/befolkning/befolkningsframskrivinger/artikler/norways-2022-national-population-projections/_/attachment/inline/37d9dfef-1cd6-4390-b6ab-1601e21b32a8:1061870b3633187b8e861856f85e2dcc6638f666/RAPP2022-28_nasjfram%20ENG.pdf.
[12] U.S. Census Bureau (2017), Methodology, Assumptions, and Inputs for the 2017 National Population Projections, https://www2.census.gov/programs-surveys/popproj/technical-documentation/methodology/methodstatement17.pdf.
[13] UN DESA (2022), World Population Prospects 2022, https://population.un.org/wpp/Publications/Files/WPP2022_Methodology.pdf.
[17] United Nations (2019), World Population Prospects 2019 - Methodology of the United Nations population estimates and projections, https://population.un.org/wpp/Publications/Files/WPP2019_Methodology.pdf.
This annex provides detail on the methods and data used in this chapter. The projection exercises are based on Eurostat population projections as well as economic and fiscal forecasts based on the OECD Long-Term Model. The latter projects economic and fiscal outcomes until 2060, following a range of inputs on productivity, employment, and population development (Guillemette and Turner, 2018[4]; 2021[15]). Production of these estimates themselves takes place in two stages. First, estimates of population dynamics are produced under different fertility scenarios and mapped to dynamics in the size of the labour force using the OECD in-house dynamic age-cohort model that, under baseline conditions, projects future labour participation by gender and five-year age-group using current rates of labour market entry and exit.
Second, estimates of public primary expenditure (i.e. government spending excluding any interest payments), potential GDP per capita and annual growth of the potential GDP per capita under each scenario are produced by combining the labour force and population forecasts with the long-term growth model presented by the OECD in the OECD Economic Outlook No. 109 (see Guillemette and Turner (2021[15]) for detail). The models estimate potential GDP and fiscal outcomes based on a range of long-term growth determinants and their long-term dynamics within the given country as well as on convergence patterns between countries across the projection period. All model inputs are detailed in Annex Table 6.A.1.
The potential economic output () in year in the model is based on a simple Cobb-Douglas production function with constant returns to scale featuring physical capital (K), trend potential employment (N) as production factors, plus trend labour efficiency (E):
Where α denotes the labour share and is fixed at 0.67. The capital and trend labour efficiency components are determined within the OECD Long-Term Model, where the latter converges to an assumed exogenous rate of global technological progress in the long-run (1.5 percent per annum) while the capital-to-output ratio stabilises in the steady-state. As such, it generally grows at the same rate as aggregate trend employment and average trend labour efficiency (see more in Guillemette and Turner (2018[4]) and Guillemette, De Mauro and Turner (2018[35])). Trend potential employment itself is obtained using the labour force participation rate forecasts of the OECD in-house labour force projection model:
Where and are the projected population and labour force participation rates for age and gender . Throughout the calculations of the baseline model and the scenarios, working age is defined as 15 to 74. The labour force is computed using aggregated population figures by gender, specifically over sex-specific 5-year age groups. The evolution of the population itself is modelled as:
Where is the mortality rate and the net migration in year for age and gender. For , new births are accounted for through the aggregate number of births for mothers at each age that survive until the end of the year, based on the age-specific fertility rate in the previous year and the total number of women in the age-specific population , also in the previous year. In each of the scenarios presented in this chapter, population dynamics are adjusted solely through differences in age-specific fertility rates. Therefore, mortality and migration rates do not change at all.
The long-term projections for public spending on pensions, health and long-term care are similarly determined within the OECD Long-Term Model. Forecasts for public pension spending depend on the ratio of retirees to workers and projected evolution of benefit ratios. It thus reflects population ageing, dynamics in the labour force and the evolution of statutory retirement ages. Projections on health and long-term care expenditure, are calibrated based on historical dynamics sourced from the OECD Health Expenditure and Financing Database. The projected growth in health and long-term care expenditure uses as input growth in GDP per capita, the population share of the elderly and the excess of healthcare price inflation over general inflation. The latter component is dependent on projected labour productivity growth to account for links between slowing productivity growth and costs pressures as well as the impact of technological change on healthcare costs (see Guillemette (2019[36]) for more information).
There are some notable limitations to using the OECD Long-Term Growth Model for this exercise. For example, the outputs of the OECD’s Long-Term Growth Model include economic production as potential economic output (potential GDP per capita), which refers to an economy at full employment of all members of the labour force. All model estimates are mechanical only and assume that any changes in fertility rates do not interact with factors outside of the model or have any indirect effects. This is especially important for employment and working hours of men and women in childbearing age, as the model does not assume any changes in employment or working time in scenarios with more or fewer births relative to baseline. That means that even with fewer births in a low fertility scenario, the model will not lead to higher employment and working time in childbearing age and may therefore underestimate the total labour input, especially of (potential) mothers.
For Norway only, substantial annual revenue from offshore oil and gas production as well as the volume and the returns to the Norwegian Government Pension Fund Global (GPFG), complicate any realistic projection of future fiscal revenues in the projections of the OECD Long-Term Growth Model. For other countries, the model assumes that government financial assets – which in the case of Norway do include the GPFG – remain constant as a share of GDP over the projection period. While this is generally in line with historical trends for most countries, for Norway it would assume that the GPFG stops growing, while Norwegian modelling expects it to decline (Guillemette and Turner, 2021[15]). For this reason, the OECD Long-Term Growth Model in its current form cannot produce reliable projections of asset and revenue dynamics for Norway. As such, this chapter is limited to an analysis of the expenditure side when considering the effects of fertility on fiscal sustainability.
Public expenditure projections on health, long-term care and pensions rely either on coefficients estimated from historical data (health and long-term care) or on stylised assumptions (pension), which account for differential spending across the age distribution. However, other public spending, including family transfers and in-kind services, is based on the assumption that governments will seek to provide a constant level of non-health/non-pension spending per capita in real terms over time, with prices for government services evolving with the wages in the rest of the economy. There is no allocation made for different levels of spending by age, so the effects of fertility rates on public spending on family benefits in cash or in kind may not be modelled fully accurate (Guillemette and Turner, 2021[15]; 2017[24]). In fact, when public health and pension expenditure is removed from historical per capita age spending profiles, the remainder of the spending is disproportionally directed towards children and young adults (Chłon-Dominczak et al., 2019[37]). As a result, the effects of fertility rates that diverge from the baseline projections will somewhat underestimate the actual effect that should be expected – which would be positive or negative depending on the specific scenario. It is therefore necessary to treat the effects on other public spending with caution.
The long-term care demand and supply are projected in a simple and illustrative setting built on a number of basic assumptions. First, it is assumed that countries would ideally want to keep the ratio between LTC workers and LTC-users constant in order to provide the same level of care as today. In effect, this acts as an upper bound on demand for care workers. or Norway, the required data on LTC workers and users is readily available and can be obtained through OECD Health at a Glance 2021 (OECD, 2021[33]). In 2019, there were 12.4 LTC workers per 100 senior citizens in Norway, a total LTC workforce of about 113 400. At the same time, there was a total of 165 200 elderly Norwegians that resided in care institutions or used home help and nursing. Taken together, this means that there were about 68 LTC workers for every 100 elderly LTC users. At the same time, the projections assume that half of all additional years of life expectancy are spent in good health, without the demand for LTC – similar to assumptions made by the European Commission (European Commission, 2021[38]). Statistics on the number of LTC users by age group – which is important to factor in the effects of increases in healthy life expectancy – are published by Statistics Norway (2022[32]).
In terms of projections on the LTC workforce, it is assumed that the share of LTC workers among the total labour force – which is sensitive to different fertility trajectories – remains unchanged in 2060. In 2019, the share of LTC workers of the labour force was slightly above 4%. The projections also factor in a scenario on productivity increases, which reduce demand for LTC workers by 20% in 2060 – about half of the productivity increases in the aggregate Norwegian economy projected in the OECD Long-Term Growth Model (see e.g. Guillemette et al., (2017[39])).
A wider discussion on the future of care services, including a projection/simulation exercise, is undertaken by the OECD in “Beyond Applause: Improving work conditions and social recognition in the long-term care sector given ageing societies” (OECD, Forthcoming[31]).This project highlights labour market imbalances in the long-term care (LTC) sector, assess their links with job quality and the quality of care in the LTC sector, and suggests ways to improve working conditions and raise social recognition in the sector.
Note: The baseline scenario refers to baseline fertility rates in the Eurostat population projections, which converge to a TFR of 1.62 by 2060 for Norway. “Low fertility” refers to fertility rates that converge to a TFR 0.5 below baseline. “High fertility” refers to fertility rates that converge to a TFR 0.5 above baseline.
Source: OECD Family Database and OECD calculations based on Eurostat Population Projections 2019 and United States 2017 National Population Projections.
Note: See note to Annex Figure 6.B.1 for a definition of the different fertility scenarios.
Source: OECD calculations based on Eurostat Population Projections 2019 and United States 2017 National Population Projections.
Note: See note to Annex Figure 6.B.1 for a definition of the different fertility scenarios.
Source: OECD calculations based on OECD in-house labour force projections, Eurostat Historical Population Projections 2019 and United States 2017 National Population Projections.
Note: See note to Annex Figure 6.B.1 for a definition of the different fertility scenarios. Primary expenditure is defined as government spending on pensions, health, long-term care and other categories (e.g. including family transfers and in-kind services), excluding any expenditure on interest payments.
Source: OECD calculations based on Eurostat Historical Population Projections 2019, United States 2017 National Population Projections and the OECD Long-Term Growth Model.