Automatic stabilisers: their extent and role.
Public sector deficits have increased markedly in nearly all OECD countries since 1990 -- the sole exceptions being Italy, Greece, the Netherlands and Portugal -- in line with the downturn in activity. Such widening has resulted mainly from cyclical movements in government revenues and expenditure rather than from specific policy actions. Cyclical changes in government borrowing tend to offset part of the fluctuation in output that would otherwise occur, hence they are often described as "automatic stabilisers". The extent of the automatic stabilisers varies across OECD countries reflecting, in particular, the overall size of the public sector, the progressivity of the tax system, the sensitivity of different tax bases to movements in economic activity, the generosity of unemployment benefit schemes and the extent to which employment fluctuates with output. This special note describes the main determinants of automatic stabilisers in OECD countries and assesses their impact on the economy.
The OECD Secretariat attempts to estimate the scale of automatic stabilisation by measuring the extent to which various forms of taxation and expenditure are sensitive to the cycle. Estimates are made of the size of government borrowing (or lending) that would occur if output were at its trend level rather than at its actual level, and thereby to estimate the cyclical component of government borrowing.
Estimates of the stabilisers cover taxation and expenditure separately. On the revenue side, all taxation receipts are adjusted, with taxes being grouped into four types. On the expenditure side, OECD Secretariat estimates of the automatic stabilisers are limited to the impact of the cycle on transfers to the unemployed, though debt-interest payments are also sensitive to some extent. While other public spending may also be cyclically sensitive (e.g. public employee wages, investment goods prices and industrial subsidies), short-term changes in these forms of expenditure require discretionary action as opposed to the automatic consequences for expenditure that stem from changes in the level of claims for unemployment benefits and related social transfers. Such a distinction is only valid in the short term as all government programmes are discretionary over a somewhat longer time horizon.
Automatic stabilisers and the tax system
The structure of the tax system has a significant impact on the size of the automatic stabilisers. The higher the average tax rate on income from a cyclically sensitive source, the larger will be the automatic stabiliser. For example, taxation is lost when an employee is made redundant. In this case, the amount of stabilisation depends on the average tax rate on labour income (defined as wage income plus social security contributions). This tax rate differs considerably between countries.(1)
The progressivity of the tax system is another significant factor in determining the size of the automatic stabilisers. On average, OECD Secretariat estimates suggest that government revenue fluctuates with a slightly greater amplitude than fluctuations in output. In part, this stems from the difference between the average to marginal rates of taxation on labour income. Such a difference means that when average income per person employed falls in a recession, either through a fall in overtime working or through a fall in wages, the drop in government revenue is more rapid than that of average incomes. In some countries (France, Italy, New Zealand and Sweden), the ratio of the average to marginal rate of taxation is close to unity.(2) More typically the ratio is around 1.4, with much higher ratios in Australia and Canada.
The sensitivity of the individual tax bases to the cycle is the final factor affecting the size of the automatic stabilisers on the revenue side. However, there are often offsetting factors at work. Employment generally fluctuates less than output over the business cycle, reflecting the hoarding of labour by employers. This limits the fluctuation of employment-related taxation at the expense TABULAR DATA OMITTED of large swings in corporate tax revenues. Such fluctuations can be particularly marked if corporations attempt to maintain dividends in periods of recession; one common method of so doing is by reducing retained profits that are often taxed at a higher rate than distributed profits. If the corporate tax rate is greater than the average tax rate on labour income, tax yields will be pushed down further in a recession.
Timing factors can sometimes delay the effect of automatic stabilisers and hence negate their stabilising role. Taxation is not always collected in the year in which the liability to pay tax accrues. For instance, income tax payments in the current tax year may be based on income in previous tax years, either for all income or only for certain types of income. Therefore, taxes, especially corporate taxes, can be related to activity in previous years. As a result of such lags, the tax system risks becoming destabilising, i.e. pro-cyclical. Payment of taxes from a boom year might only occur when the economy is already moving into recession. However, if economic agents accumulate reserves for the payment of taxation when the liability accrues, as is likely to be the case for corporations, the tax system will remain stabilising despite the lags in actual payments.
Automatic stabilisers and expenditure
Unemployment benefits are the principal source of automatic fluctuation in government expenditure. In 1991, unemployment benefits amounted to almost 3 1/2 per cent on average of total government spending and about 1 1/2 per cent of GDP. If it is assumed that expenditure on unemployment is proportional to the numbers unemployed, a 1 percentage point increase in the unemployment rate will generate an increase in government expenditure amounting to 0.2 per cent of GDP. Given the rule of thumb that unemployment changes about half as rapidly as output, a 1 per cent fall in the level of output raises government expenditure by about 0.1 per cent.(3) The impact is greater in those countries (such as Canada, Austria, Denmark, Sweden and Switzerland) where the ratio of unemployment benefit to average earnings -- the so-called "benefit replacement ratio" -- is high. Indeed, in these countries, a fall in employment can generate an increase in expenditure and reduction in tax revenues (and hence increase in borrowing) that is almost equivalent to the income from the lost employment. On the other hand, in low-tax, low-benefit countries (the United States, Japan and the United TABULAR DATA OMITTED Kingdom), the borrowing offset to fluctuations in employment income is less than one-half.
Interest payments are also cyclically sensitive. As borrowing increases, interest payments first rise and then fall as revenues recover. The scale of the fluctuation is small as interest payments in a given year are only a small percentage of any increase in borrowing. For a typical OECD country, the sensitivity of interest payments to a given fluctuation in output is about one-quarter the sensitivity of expenditure on unemployment benefits.
The impact of automatic stabilisers on government borrowing
Overall, the impact of the automatic stabilisers varies significantly between countries. The most important factor is the size of the public sector in an economy, and more particularly the importance of tax revenues. The higher the share of tax revenues in the economy, the greater is the sensitivity of government income to fluctuations in GDP(4). Thus, countries with low taxes relative to GDP (e.g. the United States, Japan, the United Kingdom and Australia) have low stabilisation from the revenue side: a 1 per cent change in GDP in these countries is associated with a change in government borrowing of about 1/3 per cent of GDP. In countries where the share of tax is relatively high (e.g. Denmark, the Netherlands, Norway and Sweden), a similar change in GDP may generate almost twice as large a change in government borrowing. Averaged over all OECD countries, a change of 1 per cent in GDP is associated with a change in borrowing of 0.5 per cent of GDP. If the weakness in demand persists, this increase becomes larger as interest payments on the original shortfall begin to mount.
Fluctuations in revenues account for a much larger share of automatic stabilisers than fluctuations in expenditure. Indeed, increases in government expenditure are only responsible for about one-fifth of the increase in government borrowing in a recession. Such a split is also reflected in a comparison of the 1992 budget plans of governments with expected outcomes. Government budgets overrun their initial targets by about 0.7 per cent of GDP in the major countries and over 1 per cent in the smaller countries. Amongst the major countries, there was little overrun of expenditure and nearly all the deficit overrun came from tax revenues that were lower than projected. The picture is more mixed for the other countries. Most had the same experience as the major countries, with the expenditure increase accounting for less than one fifth of the deficit overrun. However, several countries (Finland, Norway and Sweden) had significant expenditure overruns linked to problems in their banking sectors. Spain was an exception in that TABULAR DATA OMITTED its deficit overrun was almost entirely caused by higher spending.
The shortfall in revenue and increase in government expenditure due to the current downturn is projected by the OECD Secretariat to amount to 1 1/2 per cent of OECD GDP in 1993 -- equivalent to just under half of net government borrowing in the OECD area.(5) This cyclical component of the deficit is particularly large -- both relative to GDP and as a proportion of the total deficit -- in the Nordic countries and in Canada. In some other European countries (France, the United Kingdom, Austria, Portugal and Spain), the cyclical component is estimated to account for between one-half and one-quarter of the total deficit.
TABULAR DATA OMITTED
The impact of automatic stabilisers on cyclical fluctuations in output
The effect of automatic stabilisers on activity can be significant or almost non-existent depending on the structure of the economy. The degree of stabilisation provided depends on the same factors that influence tax and expenditure multipliers following discretionary changes in fiscal policy: trade flows, savings reactions and the degree of flexibility in labour and product markets. Simulations, using the OECD's INTERLINK model, suggest that the automatic stabilisers reduce the amplitude of cyclical fluctuations by about one-quarter in the major European countries.(6) The reduction in the cumulative output loss during a recession is about one-half. For the smaller European economies, the stabilisers are less effective in smoothing output fluctuations due to the greater exposure of these countries to foreign trade. Indeed, in some smaller countries, the stabilisers have very little impact on output. Amongst the smaller European countries, the degree of output smoothing provided by the automatic stabilisers only matches that achieved in the major European countries when the automatic stabilisers are particularly large (as in the case of Denmark, Norway and Sweden). In the United States and Japan, the stabilisers have much greater effect on the economy despite their smaller size, reflecting the below-average openness of these two economies.
Some policy issues
The difference in the size of the automatic stabilisers between OECD countries might mean that, in the face of similar deflationary forces, there could be more pressure for discretionary fiscal action in countries with low stabilisers; countries with large automatic stabilisers might conversely wish to offset some of the increases in borrowing if government borrowing and debt are already high. Moreover, the countries with significant cyclical fluctuations in general government borrowing (for a given change in output) tend to be those small European countries where INTERLINK simulations suggest that fiscal measures (automatic or discretionary) have the least impact on output. Until output returns to its trend level, there is a risk that the borrowing caused by automatic stabilisers may (through increased financing costs) turn cyclical deficits into structural deficits, especially when current output is low relative to previous trends and there is doubt about future trends in output growth and the speed of the recovery.
Tax-based automatic stabilisers have the advantage that they are rule-based. Abstracting from the timing factors mentioned earlier, they respond immediately to changes in activity and, importantly, generate expectations of future reversals that may limit the impact of greater public borrowing on long-term interest rates. Discretionary changes in expenditure that compensate for relatively low automatic stabilisers have to be carefully designed if they are not to create expectations of permanent increases in expenditure and deficits that may adversely impact on long-term interest rates. The effect of automatic stabilisers on government borrowing is more likely to be ignored by markets if they are confident, not just of reversal of the borrowing but of eventual repayment of the cyclical borrowing -- so that cyclical effects cancel out over a complete cycle. The cyclical effects on borrowing will, however, only cancel out if output lost during cyclical troughs is completely regained by above-potential output during the boom phase of the cycle. It may be difficult, though, to run the economy at a level of capacity utilisation sufficiently high to offset a marked recession without experiencing an undesirable increase in the inflation rate. If so, the automatic stabilisers would not be fully reversed in the upswing.
In addition, there is some evidence from the 1980s that increases in tax revenues stemming from the automatic stabilisers -- i.e. above-average revenues in a period of above-average growth -- were used to reduce tax rates rather than public debt. As a result, the increase in the debt ratio that occurred in the early 1980s recession was never subsequently reversed. The most that was achieved was a stabilisation of the debt ratio at the end of the 1980s, meaning that the current increase in government debt relative to GDP is taking place from a worse initial position than in the previous cycle.
1. The tables use the tax rate paid by a representative employee -- the so-called "average production worker" in the manufacturing sector. The earnings of this "typical" employee are used by the OECD Secretariat as a standard method of comparing personal taxation on employment income across countries.
2. The average tax rate on labour income is calculated as the sum of employer and employee social security contributions and income tax payments divided by the sum of salary plus employers' social security contributions. The marginal tax rate is defined as the additional amount of income tax and employees' and employers' social security contributions due to a (small) additional amount of taxable income. Taxes include state and local taxes on wage income, but exclude taxes on other types of income, e.g. capital gains or interest earnings. The average and marginal rates have been evaluated at the income level of the average production worker assuming that the spouse does not work and that there are two children in the household. The replacement rate for unemployment benefit has also been calculated as a proportion of average earnings plus employers' social security contributions.
3. Okun coefficients relating the change in unemployment to the change in output are shown in Chouraqui, J.-C., R.P. Hagemann and N. Sartor (1990), "Indicators of fiscal policy: a re-examination", OECD Economics and Statistics Department Working Papers, No. 78, Table A3.
4. The extent of the cyclical fluctuation in government revenues depends on two factors: i) the size of the initial level of taxation (the average tax rate); and ii) the elasticity of taxation with respect to changes in output (the marginal tax rate). For all OECD countries, the typical elasticity is 1.2, indicating a slightly progressive tax structure. Such an elasticity is reflected in the slope of the line relating the sensitivity of government borrowing to the average tax rate. Across countries, the size of the cyclical fluctuation in borrowing is more related to the average tax rate in the economy than to the marginal tax rate. For instance, the United Kingdom and the Netherlands have a similar tax elasticity, but the cyclical sensitivity of borrowing is greater in the Netherlands than in the United Kingdom because of the higher average tax rate in the former country.
5. In order to estimate the cyclical component of the government deficit, the OECD Secretariat first estimates trend output by using a semi-logarithmic regression of output on time. The time trends are split at cyclical peaks. For periods after the last cyclical peak (generally in 1990 or 1991), the trends are projected judgementally taking into account labour force growth, capital formation and technical progress. The estimate of trend output enables the cyclical shortfall in output to be determined. This, in turn, generates an estimate of the shortfall in each type of revenue and expenditure, given the sensitivity of each category to output fluctuations.
6. Private consumption was forced to move in such a way as to generate a downturn and subsequent upturn in economic activity. The downturn was assumed to have a duration of three years, with a maximum fall of 1 per cent of baseline output. In the simulation that determined the size of the necessary changes to saving to achieve this fluctuation, the automatic stabilisers were allowed to work. In the second stage, these changes to saving were imposed on the baseline but governments were assumed to change taxation in order to prevent any increase in the government deficit. The difference between the output loss in the second case and the initial imposed loss represents the impact of the automatic stabilisers. In both simulations interest rates were held constant and exchange rates were allowed to float.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||public sector deficits of Organization for Economic Cooperation and Development countries|
|Publication:||OECD Economic Outlook|
|Date:||Jun 1, 1993|
|Previous Article:||The turmoil in European currency markets.|
|Next Article:||Rising external imbalances: a comparison with the mid-1980s.|