Can an ageing Scotland afford independence?
Keywords: Scotland; independence; OLG; government spending; population ageing
JEL Classifications: C68; E17; H53; J11
In the light of the current Scottish independence debate, much attention is being paid to whether Scotland and the rest of the UK will be better off after the separation. One of the arguments that has been raised during the debate is that Scotland is in a worse demographic situation than the rest of the UK, and independence will make the fiscal challenge harder in the context of its ageing population. In 2012, the old-age dependency ratio (OADR) (1) in Scotland and the rest of the UK was the same at 29 per cent. However, in the future the Scottish population is projected to age more rapidly, and by 2037 OADR in Scotland will reach 48 per cent, while in RUK it will reach 45 per cent. (2) However, by 2050 OADR in both regions will converge again.
This paper uses a dynamic multiregional Overlapping Generations Computable General Equilibrium (OLG-- CGE) model for Scotland, the rest of the UK (RUK) and the rest of the World (ROW) to evaluate demographic scenarios for Scotland and RUK with and without independence. The model is in the Auerbach and Kotlikoff (1987) tradition and introduces age-specific mortality following Boersch-Supan et al. (2006).
We use the 2012-based principal population projections produced by the Office for National Statistics (ONS) as an exogenous demographic shock for Scotland and RUK. The status quo scenario assumes that Scotland stays part of the UK and all government expenditures associated with ageing population (mainly pensions and health) are funded on a UK-wide basis. In the status quo scenario, population ageing has a strong impact on economic development in both regions. By 2060 output per person falls in Scotland and RUK by 10 per cent and total government spending increases by 4 percentage points of GDP. To achieve government budget balance the effective labour income tax rate has to increase from about 13.0 per cent to 21.5 per cent, i.e. by 8.5 percentage points.
In the independence scenario, RUK and Scotland have separate government budget constraints and each of them has to pay for the growing demands of their ageing populations independently. As expected, Scotland is worse off in the case of independence, although the difference with the status quo scenario is not large, especially when compared with the overall effect of the demographic shock. The greatest difference is in 2035 and at this point the effective labour income tax rate in Scotland is 1.4 percentage points higher in the independence scenario compared with the status quo scenario. The difference is non-negligible but it is much smaller than the effect of population ageing itself.
It is interesting to compare the results presented in this paper with the Institute for Fiscal Studies (IFS) report on the fiscal sustainability of an independent Scotland (Amior et al., 2013) and Crawford and Tetlow (2014) because they address some of the same issues. Our conclusions differ somewhat for a number of reasons. First and most importantly, we set ourselves different goals. This paper looks only at the impact of demographic change in the context of potential Scottish independence, while the IFS team tries to take into account all the factors that might influence the Scottish fiscal position. One of the very important consequences of this is the different treatment of future North Sea revenues. Second, we are using very different models; our model has a general equilibrium structure with three interacting regions and takes into account behavioural and general equilibrium effects that tend to benefit Scotland, while the IFS model concentrates only on the fiscal sector (with much greater detail) and relies on assumptions about the rest of the UK and world economy. Third, we use different population projections for our central simulations; this paper uses the latest 2012-based ONS principal projections, while the IFS uses the low-migration scenario of the 2010-based population projections. The choice of demographic assumptions is important and 2010-based projections are more negative for Scotland, which is discussed further in section 3.1. But despite the differences between our approaches our main conclusion is supported by Crawford and Tetlow (2014) "... our model suggests that, while population ageing will increase public spending quite substantially in both Scotland and the UK as a whole, the difference between the size of this effect in Scotland and the UK will be quite small."
2. The model
The model economy is made up of three regions: Scotland, the rest of the UK (RUK) and the rest of the World (ROW), which closes the model and combines North America, Europe (EU-15), China, India and Japan. Household behaviour is captured by 21 representative households that interact in an Allais-Samuelson overlapping generations structure representing 5-year age groups (0-4, 5-9, ..., 95-99, 100+). Lifetime is uncertain and unintended bequests are distributed through a perfect annuity market, as described theoretically by Yaari (1965) and implemented in an OLG context by Boersch--Supan et al. (2006). Each region in the model produces one internationally traded good which is an imperfect substitute for the goods produced in other regions (Armington, 1969). With perfect foresight, selfish and optimising individuals in each region consume a basket of all the imperfectly substitutable goods. Firms maximise profits, hire labour and rent physical capital. The accumulation of each region's capital stock is given by the usual law of motion subject to depreciation. There is perfect substitution between domestic assets (physical assets and government bonds), and perfect financial capital mobility across regions. The government budget constraint includes spending on health and education, which is assumed to be constant in real terms per person of specific age. Other government spending is assumed to be age-independent. The pension programme is a part of the overall government budget and public pensions are indexed to wages. The model imposes the condition that governments target a constant debt-to-GDP ratio by setting the effective labour income tax rate every period. Other tax rates (i.e. on capital and consumption) remain constant.
The equilibrium condition for the goods market is that each regional output must be equal to total demand originating from all regions as well as internal demand for health and education services provided by the government. Labour and physical capital are immobile across regions and a market exists for these two factors in each region. The world stock of wealth accumulated at the end of a given period must be equal to the value of the world stock of government bonds and stock of physical capital. A full description of the model with complete equations listing is provided in Lisenkova and Merette (2013).
Scotland and RUK as one country
The only differences in the status quo scenario when Scotland stays part of the UK are within the government sector. Scotland and RUK have joint government budget constraints and all tax rates and public debt-to-GDP ratio are equal between the UK regions.
The Scotland and RUK models are calibrated using 2010 data. The data on public finances, GDP components and international trade are taken from the ONS, HM Treasury and Scottish government. Effective labour income, consumption and capital tax rates are calculated from the corresponding government revenue categories and calibrated tax bases. Data on total amount of pensions are taken from the Government Actuarial Department (GAD). Based on this information, the effective pension contribution rate and the average size of pension benefits can be obtained. For simplicity, it is assumed that both males and females start receiving pension benefits at the age of 65, i.e. we do not take into account further planned increases in the state pension age. The source of the labour market data is the Labour Force Survey (LFS).
The estimates of the age structure of government spending on health and education are taken from the UK National Transfer Accounts for 2007, constructed by McCarthy and Sefton (2010). There is a growing literature that argues that health spending depends largely on proximity to death (Zweifel et al., 1999; 2004; Gray, 2005). To account for increasing life expectancy, we extrapolate 2007 health spending profiles for future years. The key assumption that we make is that a one year increase in life expectancy corresponds to a one year increase in healthy life expectancy. Starting from the age of 50, a one year improvement in longevity results in a one year shift back in the health profile.
The main source of data for ROW is the GTAP-8 database (Narayanan et al., 2012). The rest of the parameters are either omitted (e.g., public consumption is not disaggregated into categories) or calibrated to give plausible results (e.g., tax rates).
For Scotland and RUK we use 2012-based ONS principal population projections. For ROW we use the 2012 revision of the UN medium variant population projections (United Nations Population Division, 2013).
Figure 1 shows that the whole world is on the verge of significant acceleration of the ageing process. ROW is much younger than both Scotland and RUK, but the speed of change in population structure is faster in ROW and by the end of the century the gap between OADR in the UK and ROW is noticeably smaller. Scotland was younger than RUK until the early 2000s, but will age more rapidly for several decades. However, by the end of the projection period OADR in both regions converges.
All simulations use 2010 as a base year. This means that the first period where shock is applied is 2015. This is just one year before the proposed year of potential separation and is unavoidable, because one period in the model lasts five years. We start with the status quo scenario, in which Scotland stays within the UK. In this case public spending is financed on a UK-wide basis. All results are shown as a percentage change relative to the baseline without population ageing unless stated otherwise.
Figure 2 demonstrates the effect of demographic change in Scotland and RUK on supply of factors of production and the aggregate level of real GDP. Over the next five decades, effective labour supply in RUK increases by 12 per cent, while in Scotland it is mostly flat. The difference in labour supply between RUK and Scotland is larger than suggested by the OADR (see figure 1). That is because in RUK both the 65+ age group (numerator) and the 20-64 age group (denominator) are increasing more rapidly than in Scotland.
Capital supply increases faster than labour supply in both regions, resulting in an increase in the capital-labour ratio. Capital accumulation is determined by saving behaviour. In this model the only motive for saving is for provision in old age. Consequently, a more rapidly ageing population is predicted to save a larger share of its output to provide for the faster growing retired population. However, this does not mean that the overall level of saving would be higher in an ageing society as lower labour supply might result in lower level of output.
By the end of the simulation period capital supply in RUK increases by 24 per cent and in Scotland by 13 per cent. As a result of these changes in supply of factors of production, by 2060 the real GDP in RUK increases by 16 per cent and in Scotland by 4 per cent.
Population ageing is characterised by a faster growth of the retired population relative to the working age population. That is why overall growth of output does not guarantee an increase in the level of output per person. Figure 3 demonstrates that changes in the age structure of the population lead to the decrease of real GDP per person in RUK and Scotland by 10 and 9 per cent respectively. Although Scotland is ageing more rapidly than RUK, faster growth in the capital-labour ratio results in a marginally slower reduction in output per person.
The dynamics of different categories of government spending expressed as a per cent of GDP is presented in figure 4. By 2060 spending on pensions increases by about 3 percentage points of GDP. Spending on health increases by about 2 percentage points of GDP while spending on education marginally decreases. Finally, age-independent government spending increases by about 0.5 percentage points of GDP. Spending on health and pensions, which are greatly affected by older age groups, grows faster in Scotland. Spending on education, which depends mostly on young age groups, declines faster in Scotland. The overall level of public spending increases by about 4 percentage points of GDP, again slightly higher in Scotland than in RUK.
The policy instrument that allows the government sector to balance its budget in this model is the effective labour income tax rate. It is a non-distortional tax as labour supply is exogenous. To keep the government budget in balance in the status quo scenario, the effective rate of labour income tax increases from about 13.0 per cent in 2010 to 21.5 per cent in 2060. The tax toll arising from the fiscal pressure of ageing is thus equal to 8.5 percentage points.
To avoid increasing the effective labour income tax rate by 2060 total factor productivity (TFP) growth has to be about 0.6 per cent per year. According to Harris and Moffat (2013), between 1997 and 2008 TFP growth in Great Britain was on average equal to 1.6 per cent per year. This means that covering the public expenditures associated with ageing--with no increase in quality or quantity of public services provided--would require over one third of recently observed TFP growth.
With some market power given by the Armington assumption and a relative production decline due to ageing, the price of the goods produced by Scotland increases relative to the price of the goods produced in RUK and ROW. In contrast, the price index of the goods consumed by households in Scotland changes less, as consumption consists of domestic and imported goods and the price of those goods moves in the opposite direction. Consequently, Scotland's terms of trade improve over time (figure 5).
As shown in figure 6, the capital-labour ratio increases more rapidly in Scotland than in RUK. This is mainly due to the impact of population ageing on labour supply as it directly affects the denominator of the capital-labour ratio. As reported above, labour supply remains flat in Scotland between 2010 and 2060 while it increases by about 12 per cent in RUK for the same period. The increase in capital-labour ratio puts downward pressure on return to capital and leads to capital outflow from the more rapidly ageing region. However, improvement in the Scottish terms of trade, discussed above, prevents large-scale capital outflow from Scotland.
Different patterns in labour supply growth between Scotland and RUK are also reflected in the effect on real wages. Due to faster population ageing in the UK compared with ROW there is growth in real wages in both UK regions. However, in Scotland real wages are growing faster than in RUK; at the peak in 2035 real wages are 9 per cent higher in RUK and 6 per cent higher in Scotland and this difference persists until the end of the simulation period.
The main question of this paper is what will happen if Scotland becomes independent and has to pay for ageing related expenditures itself. Figure 7 shows the percentage point difference in the effective labour income tax rate between the independence and the status quo scenarios. The highest difference is in 2035 and at this point the labour income tax rate in Scotland is 1.4 percentage points higher in the independence scenario. This effect is non-negligible. However, it is much smaller than the effect of population ageing discussed above--8.5 percentage points by 2060 in the status quo scenario.
The results presented in the previous section are sensitive to the demographic assumptions. To analyse sensitivity to population projections we compare main results with simulations based on the low migration variant of the 2010-based ONS population projections. These are the same projections that were used by the Institute for Fiscal Studies in their recent report on the fiscal sustainability of an independent Scotland (Amior et al., 2013) and by Crawford and Tetlow (2014). Figure 8 shows OADR for both regions for 2010-based low migration projections. Comparing this with figure 1 reveals a very different picture. By 2060 in the 2010-based low migration projections RUK has lower OADR than in 2012-based principal projections, 50 per cent compared with 51 per cent, while Scotland has higher OADR, 56 per cent compared with 52 per cent. (3)
The relatively better situation for RUK in the 2010-based low migration population projections makes the overall effect of ageing at the UK-wide level less severe--for the status quo scenario the effective labour income tax rate has to increase by 7 percentage points, compared with 8.5 percentage points in our main results. At the same time the difference between the two scenarios more than doubles compared to our main results, reaching 3 percentage points compared with 1.4 percentage points. Thus, using the population projections which are less favourable for Scotland results in the additional cost of independence being almost half as large as the costs of ageing itself. For comparison, in the case of our main simulations it only constitutes 16 per cent.
The high sensitivity of the results to the demographic projections suggests that active policy aimed at improving the demographic situation might play an important role in reducing the effects of population ageing in both regions. For example, if Scotland wants to reach the same OADR as RUK in 2060, it has to increase its net migration by roughly 50 per cent compared with the principal ONS migration assumption. This would amount to approximately 23,000 per year, which is equal to the average level of net migration that Scotland experienced between 2004 and 2011, and is below the long-term net migration assumption in the high migration variant by the ONS (24,000 per year). Thus, our conclusions confirm the claims of the current Scottish government that lack of control over Scottish immigration policy is one of the important arguments for the case of Scottish independence.
Population ageing over the next 50 years will be a major demographic challenge in many regions of the world, including the UK. To investigate the case of the UK, with and without Scottish independence, we developed a multi-regional OLG-CGE model for Scotland, the rest of the UK (RUK) and the rest of the World (ROW).
To study the interaction of population ageing and potential Scottish independence, we conducted simulations under two situations: the status quo and the independence scenarios. According to the simulations presented in this paper, population ageing has a substantial economic impact on Scotland and RUK in both scenarios. In the status quo scenario, by 2060, output per person falls in Scotland and RUK by 9 per cent and 10 per cent respectively and total government spending increases by about 4 percentage points of GDP in both regions. These numbers suggest that quantity and/or quality of public services in the UK are likely to decline unless productivity growth compensates for population ageing. Indeed, under the assumption that the government maintains a balanced budget, the effective labour income tax rate needs to increase from 13 per cent to 21.5 per cent. Alternatively, to avoid increasing the level of taxation, keeping a balanced budget requires total factor productivity growth of 0.6 per cent per year.
Scotland is worse off in the case of independence as against the status quo. The effective labour income tax rate in the independence scenario has to increase further compared with the status quo scenario. The additional increase reaches its maximum in 2035 at 1.4 percentage points. The additional rise in the tax rate is non-negligible, but is much smaller than the population ageing effect (status quo scenario) which generates an increase of about 8.5 percentage points by 2060.
The low-migration variant of 2010-based ONS population projections would more than double the cost of independence. But an independent Scotland will have the power to set immigration policy in order to tackle its demographic challenges.
The aim and scope of this paper is to isolate the effects of population ageing in the context of potential Scottish independence. However, other factors may potentially make the case for independence less appealing, such as a decline in the revenues from the North Sea and the level of public debt in an independent Scotland.
The bottom line is that, clearly, population ageing is a major issue for Scotland and RUK, regardless of the final result of the independence vote. But unless the speed and intensity of population ageing in Scotland increases rapidly relative to RUK in the years to come, demographic change is not a strong argument influencing the choice between the status quo and independence.
Amior, M., Crawford, R. and Tetlow, G. (2013), Fiscal Sustainability of an Independent Scotland, IFS report.
Armington, P.S. (1969), 'A theory of demand for products distinguished by place of production', IMF Staff Papers, 16, 159-76.
Auerbach, A. and Kotlikoff, L. (1987), Dynamic Fiscal Policy, Cambridge, Cambridge University Press.
Boersch-Supan, A., Ludwig, A. and Winter, J. (2006), 'Aging, pension reform and capital flows: a multi-country simulation model', Economica, 73, pp. 625-58.
Crawford, R. and Tetlow, G. (2014), 'Fiscal challenges and opportunities for an independent Scotland', National Institute Economic Review, 227.
Gray, A. (2005), 'Population ageing and health care expenditure', Ageing Horizons, 2, pp. 15-20.
Harris, R. and Moffat, J. (2013), 'The direct contribution of FDI to productivity growth in Britain, 1997-2008', The World Economy, 36(6), pp. 713-36.
Lisenkova, K. and Merette, M. (2013), 'Can an ageing Scotland afford independence?', NIESR Discussion Paper no. 418, London, National Institute of Economic and Social Research.
McCarthy, D. and Sefton, J. (2010), First Estimates of UK National Transfer Accounts.
Narayanan, B., Aguiar, A. and McDougall, R. (Eds) (2012), Global Trade, Assistance, and Production: The GTAP 8 Data Base, Center for Global Trade Analysis, Purdue University.
United Nations Population Division (2013), World Population Prospects: The 2012 Revision.
Yaari, M.E. (1965), 'Uncertain lifetime, life insurance, and the theory of the consumer', Review of Economic Studies, 32, pp. 137-60.
Zweifel, P., Felder, S. and Meiers, M. (1999), 'Ageing of population and health care expenditure: a red herring?', Health Economics, 8, pp. 485-96.
Zweifel, P., Felder, S. and Werblow, A. (2004), 'Population ageing and health care expenditure: new evidence on the "Red Herring'", The Geneva Papers on Risk and Insurance, 29, 4, pp. 652-66.
(1) Here defined as population aged 65+ divided by population aged 20-64.
(2) According to the 2012-based ONS principal projection.
(3) OADR in RUK is almost the same in the low-migration scenario of 2010-based projections compared with principal 2012-based projections because they have almost the same long-term migration assumptions: 131,000 and 149,500 respectively (12 per cent difference). The difference is even smaller if we look at the migration assumption for 2012-16 (7 per cent). At the same time, Scottish long-term migration assumptions are much lower in the low migration scenario of 2010-based projections compared with the principal scenario of 2012-based projections: 9000 and 15,500 respectively (42 per cent difference). Other components of population change (fertility and mortality) as well as population age structure also play a role.
Katerina Lisenkova * and Marcel Merette**
* National Institute of Economic and Social Research and Centre for Macroeconomics. E-mail: firstname.lastname@example.org. ** University of Ottowa. Financial support from the Economic and Social Research Council under the grants 'A dynamic multiregional OLG-CGE model for the study of population ageing in the UK' and 'Scottish Centre on Constitutional Change' is gratefully acknowledged.
|Printer friendly Cite/link Email Feedback|
|Author:||Lisenkova, Katerina; Merette, Marcel|
|Publication:||National Institute Economic Review|
|Date:||Feb 1, 2014|
|Previous Article:||Funding pensions in Scotland: would independence matter?|
|Next Article:||Fiscal challenges and opportunities for an independent Scotland.|