I. An Overview: high government spending, does this reflect a European preference for intervention? Greater burden than in the rest of the OECD ...

Government spending comprises on the one hand the reflection of choices in government policy and the degree of State intervention in the economy: the Anglo-Saxon liberal model is therefore opposed to the continental European model which is more interventionist by nature. Compared with the other OECD countries the countries of the EU set themselves apart by the burden of their government spending: on average this represented 48% of the GDP in 2002 versus 45% on average within the OECD, 35% in the USA and 38% in Japan.

...even if the social service systems are included.

It is now a known fact that rough comparisons must take into account the organisation of the social services systems: on the one hand private funding of a major part of social services influences the structure and the amount of government spending, whilst the use of private insurance is obligatory in some cases. On the other hand the gross data of government spending does not reveal the amount of social work undertaken by a country, because fiscal measures with a social vocation are simply not taken into account.

OECD studies show that if these factors are included the standard deviation in expenditure ratios in the GDP between States is reduced by more than 40% in comparison with the existing deviations for gross public spending data. A consideration such as this also leads to a different ranking of countries in terms of the social expenditure ratio/GDP: in 2002 the USA was the 3rd country with the lowest gross level of expenditure/GDP, but 10th in the group of 18 studied, if we consider all net social expenditure[5] brought to the GDP.

Nevertheless if some EU members lie below the OECD average (UK notably for the gross rate, Ireland and the Czech Republic in particular if we look at social expenditure) and if the difference between the other OECD countries is narrowed down most EU members lie high in the margin and those countries with the highest ratio of public expenditure in the GDP are all EU members (Sweden, Denmark, France in the previously mentioned 2002 ranking for gross levels, Sweden, Germany and Belgium for the burden of social spending [6]).

A comparable level of public spending in the new Member States in comparison with EU15

A year and a half after the entry of ten new Member States into the EU we might ask what impact has been made by their spending structure on that of the EU: is it drawing the EU closer or further away from the rest of the OECD?

On the perimeter of EU15 the burden of public spending in 2004 was 46.8% of the GDP. In EU25 it lies at 47.9%. The burden of public expenditure in EU25 is still lower than public expenditure in the Euro Zone countries[7].

The entry of the new Member States has therefore slightly increased the burden of public expenditure in the European Union but to a lesser degree. Above all this can be explained by the new members' low GDP in comparison with that of EU15.

Beyond their situation in terms of budgetary balance – all in deficit in 2004 except for Estonia and beyond 3% of the GDP in five countries including Poland and Hungary we might also ask what each of these countries' level of public spending is since they have been undergoing an economic transition process for nearly fifteen years now.

The burden of public spending varies between 34.3% in Lithuania and 54.1% in Malta. We can distinguish five countries with public spending higher or equal to that of the 15 (Malta, Poland, Hungary, Slovenia and the Czech Republic): these are States that are amongst those with the highest GDP. Conversely Slovakia, Cyprus and the Baltic countries are clearly below the average.

The "Wagner Law" that says that the level of socialisation of activities depends on the level of development remains therefore unproven if we compare the new Member States with the 15: the transition process has not removed evidence of the old administrated and mostly government controlled economy; the slightest GDP rate per head does not equate with a lesser burden in public spending for the ten countries studied. We can see however that in countries such as Cyprus or Slovenia whose GDP per capita is closest to that noted in the Euro Zone (76% in Cyprus and 72% in Slovenia in 2003[8]), public spending is higher than in the Baltic countries whose GDP per capita is lower (between 40 and 45% of the GDP per capita in the Euro Zone in 2003).

II. Finding room for manœuvre once more – a necessity for all, contrasted situations.

The analysis of problems with regard to the sustainability of public spending comes from the generalised growth of debt in the 1980's and the beginning of the 1990's within the OECD countries: for public finance to be considered sustainable there must be a debt growth level that does not rise durably beyond the growth of the nominal interest rate due to the debt.

Sustainability therefore raises two issues with regard to public funding finance:

Several factors endanger the sustainability of spending.

Work undertaken by the OECD and also by the EU largely blame the impact of the ageing population on public spending: the commission for the European Economic Policy showed in October 2003 if policies remained the same, public spending linked to ageing: pensions, healthcare expenditure, dependency expenditure, might represent in 2050 3 to 9 additional points of the percentage of the GDP in most Member States (up to 13 points in Greece). In France the rise was estimated at 4% of the GDP ie around 60 billion € of additional annual expenses in conjunction with the 26.5 billion € in 2000.

In some Union States as in the new members but also in Germany in the new Länder or in the UK heavy public investment requirements have also been identified and must be financed.

Not all the Union countries are faced with the same situation

Work undertaken by the Commission on the economic perspectives of the Member States now regularly include updated forecasts on the development of expenditure linked to ageing as far ahead as 2050 if possible – not all of the countries undertake such distant forecasts however (hence France stops at 2040). They also try to measure commitments off the balance sheet or attempt to establish the level of implicit public debt.

In its 2005 report on public finance the European Commission included an overall appraisal of the sustainability of public funding of each of the fifteen Eu Member States before the enlargement of 2004.

By integrating a series of criteria that are both qualitative and quantitative[9], three appraisal levels have been defined:

The degree of sustainability does not depend on the level of public expenditure, as illustrated in the analysis results published by the Commission:

We note that those countries whose public spending is lower than the EU average are confronted by perspectives deemed as sustainable.

But undoubtedly it is more interesting to note that four of the five countries that have the greatest expenditure also have public finance that is believed to be sustainable. Indeed the control of public finance is frequently presented as being linked to a low level of public spending; the study of the situation in EU25 reveals however more complex models.

This leads us to look at the control some Member States have over their public finance whatever their level of expenditure in the shorter term.

Other surprising results in the development of expenditure and public funding in the short term...

If we look at developments in the burden of public expenditure between 2003 and 2006 –taking for the latter the forecasts undertaken by the Commission in its 2005 report as a reference – we see extremely varied situations which are apparently without correlation with the "starting point", ie with the burden of public spending in the GDP. The graph below ranks the countries according to the development of public spending in absolute values:

A first group of countries, typified by a high level of public spending witnesses little development in this over the 2003-2006 period (development lower or equal to 0.6% in absolute value): these countries are Portugal, France and the Netherlands[10]. Ireland whose level of public spending is lower than the other three can be ranked in this group.

A second group of countries witnesses a development in expenditure between 1 and 2% (ie around the average of EU25) : these are Malta, Slovakia, Italy, Greece, the UK, Poland, Lithuania, Spain, Finland, Luxembourg and Slovenia. This group is heterogeneous from the point of view of the burden of public spending.

A third group of countries (if we exclude Latvia) has decreased the burden of its public spending by over 3% (up to 5.3%) between 2003 and 2006, although the level of the latter is high: these countries include Sweden, Denmark, Belgium, Austria and Germany.

Finally four countries from the last enlargement (Hungary, Estonia, Cyprus and the Czech Republic) have witnessed more sizeable developments, in terms of a decrease (except for Estonia).

The third group shows that countries whose degree of economic socialisation is high are not necessarily the countries where expenditure is extremely rigid.

This primary empirical observation might be completed by the study of the budgetary balances of EU25, in comparison with their level of public spending. The graph below shows the Member States by ratio of public spending/GDP and indicates in each case the gross therefore "spontaneous" budgetary balance (in order to obtain data for the new Member States).

Amongst the seven countries whose budgetary balance is positive or zero, four have a level of public spending higher than the EU average (Sweden, Denmark, Finland and Belgium). Three of these countries were part of the group whose public spending burden decreased by more than 3% over the period 2003-2006 and Finland was close to it, reaching a reduction of 2%. We might also add Austria to this group whose budgetary deficit was limited to 1.3% in 2004.

Overall amongst the twelve countries of the EU that lie beyond the EU25 average in terms of public spending burden, five have a budgetary deficit of 3% of which only one in the Fifteen (France). If we only look at the Fifteen, out of the ten Member States that have higher public spending than the average of the Fifteen, three go beyond the 3% deficit threshold, versus one amongst the five States who lie below the average (UK).

More generally the graphs and tables above show that there is no absolute correlation between the state of public finance and the level of public spending.

A relatively large group of countries at present combine:

This applies notably to the Scandinavian countries (Denmark, Sweden and Finland), Austria and to a lesser degree Belgium.

A contrario, the situation of States where a budgetary deficit has accumulated beyond the Stability and Growth Pact threshold (in 2004 9 countries had a deficit higher than 3%) high and apparently rigid public spending along with the anticipation of long term sustainability problems seem to be of concern. This applies notably to France and Greece who are accumulating handicaps.

Finally on the perimeter of the Fifteen an intermediary group reveals a contrasted situation either because budgetary control is uncertain and must face up to major expenditure requirements (the UK for example) or because the strategy to control spending has not yet been conclusive (Germany or Italy for example), or that public spending is apparently more rigid (Portugal).

Using these observations as a base it is possible to examine some strategies of public spending control transversally.

III. How do we win back room for manœuvre for expenditure?

As an introduction we must recall that from one country to another, governments do not have the same levers over the entire perimeter of public spending, notably on local government or community spending and on social protection spending.

Schematically we can distinguish three types of approach in the control of spending: a quantitative approach that aims to establish expenditure development norms, a "structural" approach, that aims to act on the source of expenditure to reduce it and a process-based approach that tries to govern expenditure by public policy priorities in the process of authorising expenditure (which might result in quantitative and/or structural approaches to control).

The quantitative approaches to limiting expenditure and behaviour rules

Using summaries established by country the OECD undertook a study that was published in the Revue économique[11] that follows an appraisal of the techniques employed.

A restrictive framework of spending exists in the Czech Republic, Denmark, Finland, Hungary, Ireland, Poland, Spain, Portugal and Sweden ie the countries that reveal, with the exception of the new Member States quite a healthy budgetary situation or in any case one that is improving (Portugal). This is incidentally quite a widely spread instrument employed across the OECD.

Finland and Ireland have adopted for example a top down budgetary approach that defines the overall budget ex ante before moving on to successive arbitration to distribute it.

A great number of countries have also improved the coherence between budgetary strategy of regional communities and national objectives (Czech Republic, Denmark, Germany, Hungary, Italy, Poland, Portugal) or have established rules on the "distribution of spending" (in Belgium, authorised deficits are established for the federal government and the social security on the one hand, for the regions and towns on the other), thereby introducing "public solidarity" into the quantitative approach.

In France there is an expenditure development norm just for the State budget, defined in real terms since the preparation of the Finance Law for 1998 and that was the subject of close monitoring in application[12].

In other countries, deficit objectives and/or tax ceilings introduce pressure on public spending. In Denmark the aim of freezing the level of fiscal pressure led to the introduction of obligatory budgetary balance for regional communities, a commitment to maintain budgetary excess on the part of central administration in the mid term and as a result a limit on expenditure.

Finally in nearly all EU Member States a more or less restrictive stabilisation of expenditure has been established or imposed over several years in coherence with the stability and growth programmes put forward within the framework of economic and monetary co-operation.

What are the advantages and disadvantages of such strategies?

A quantitative approach enables action on expenditure independent of the development in revenues and facilitates a discussion focussed on objectives and requirements rather than on the possible "distribution of profits" (on condition that expenditure objectives are established at growth level lower than the GDP, which is generally the case).

Then work must be done in order to respect the spending limits thereby revealing the burden of restricted expenditure which must theoretically encourage the awareness of resulting or future rigidity.

If these are monitored they enable action on the budgetary situation and therefore on the dynamics of debt, in the short term at least.

They reveal however the dangers of a short term approach (concentration on the development of short term expenditure) leading to an eviction of discretionary spending such as the investment in infrastructures or research; indeed it is easier to act on this spending than on the service of the debt or on civil servants' pay, when there is guaranteed employment for civil servants (which is the case in almost all of the EU countries). The pronounced contraction of public investment in EU countries over the period of public finance stabilisation during the 1990's illustrates this phenomenon and highlighted by the European Commission (2003). The UK, in the 1980's and 1990's notoriously under invested in public services in order to alleviate fiscal pressure, giving rise today to the return of an even keel. The quantitative approach can therefore result in "bad spending driving out the good spending" ie spending in capital that enables both the stimulation of growth potential and as a result the improvement of finance perspectives for public spending long term.

Several of the Union's countries (Germany and the UK notably) have introduced "golden rules" that limit deficit (Germany) or borrowing (UK) to the level of investment, in order to target more restrictive spending reductions. There is however major debate on the pertinence and effectiveness of such rules, notably given the difficulty in defining the perimeter of investment: teachers' salaries that are going to increase the human capital of their pupils are not included, whilst all the public infrastructures are part of it, including those whose economic impact has not been proven (a frequently quoted example is public swimming pools!).

Although the approach remains purely quantitative it runs the risk of having limited effects on the sustainability of public finance and that is without mentioning its sensitivity to the spontaneous development of expenditure due to the position of the economic cycle.

Spending norms can incidentally be diverted: the main techniques used that have tended to develop across the Member States recently are fiscal spending and debudgetting: fiscal spending (which the Anglo-Saxon countries and notably the USA have employed for a long time) eat away at revenues, without easily affecting the level of public spending. Sometimes it is justified due to its greater effectiveness in economic terms in comparison with direct intervention expenditure[13]. In other cases they comprise a deviation of the expenditure norm.

Debudgetting follows the same kind of logic. It can hardly be justified apart from the desire to isolate the funding of some expenditure by allocated revenues. It helps break down public spending into sub-sectors on which it is much more difficult to act. PFIs (Private Finance Initiatives), which notably target operational efficiency profits, have comparable results since these funding techniques for public investment enable a smoothing of the budgetary burden over time by transferring a part of the risk over to the private partner.

Finally a quantitative approach born most often of the desire to improve the budgetary balance can paradoxically slow the reduction in rigid spending by means of structural reforms Indeed the expected advantage in terms of controlling spending most often follows a J curve (a reform that is initially expensive [14], requiring savings or mid-term spending control). In its 2005 report on public funding the Commission tried to assess the links between structural reforms[15] and the budgetary situation of the Member States: this did not reveal less frequent structural reforms in years of budgetary consolidation.

If this is the case then the analysis of the debates in EU countries notably on the occasion of the revision of the Stability and Growth Pact clearly demonstrate that budgetary restrictions are seen as an encouragement to focus on the short term. The Commission also observes in its report that there is no significant difference between the primary corrected budgetary balance and the variations in the cycle in the years that precede and follow the application of a structural reform given the reforms that have been analysed. It is important a minima that the quantitative approach is defined over a period of several years making it possible to compensate for the effects of the cycle.

Structural Reforms

Most of the 25 Union members have established so-called structural reforms in the sense that they durably modify the running of the economy in its public dimension at least, (for example the conditions to create prices on markets). They generally have the dual objective of improving the running of the economy (stimulation of growth) and reducing public deficit. Germany recently provided the quite comprehensive example of a policy of structural reforms aiming to reduce public spending ... and the difficulties experienced in "assuming" such reforms politically without the effects being visible other than on the state of public finance.

To study their effect on spending we might take as an example the reforms established in all of the Union countries in terms of retirement regimes, that have had a structural effect both on the economy (running of the employment market) and on public spending: summarised appraisals by the Commission in 2004 show reductions in most Member States in the increasing burden of expenditure on pensions in the GDP anticipated long term, in the wake of ongoing or completed reforms, in comparison with appraisals undertaken in 2000 (cf. supra).

These beneficial effects are however identified long term, and moreover according to approximate methods that have strong links to underlying hypotheses relative to the growth of the economy or the development of the level of employment and unemployment.

Even if we do not take these objections into account it remains that the reforms have differentiated effects depending on whether the reform has significantly changed the retirement regimes or not (either by modifying the parameters of the retirement regimes via distribution or by completing these regimes by public or private funds): a "painful" reform does not necessarily settle the problems of the sustainability of public spending. Hence forecasts undertaken by the Commission on the development of public spending up to 2050 show that in most countries the most notable increase still involves spending on retirement, whose burden in the GDP increases everywhere except for in Poland and Estonia but in very variable proportions, ranging from between +0.8 points of the GDP in Sweden to 10.3 points in Greece.

The Scandinavian countries (Sweden, Finland and Denmark) chose a long time ago to place liquid assets in public reserve funds in order to capitalise revenues and face up to future pensions enabling them to control expenditure.

Structural reforms that enable durable action on public spending present the risk of a delayed impact on spending, which is sometimes uncertain and endured politically if they are not based, as in the Scandinavian countries, on a relatively long term political and social consensus.

The "process-based" approach

Unlike the methods illustrated above, the approach defined by default here as "process-based" places the control of public spending on enlightened choices made during the budgetary process and not on an objective "per se" to reduce public spending.

According to work undertaken by the OECD nearly all Members have established within the budgetary process measures that enable concentration on overall priorities, on the results of programmes and political assessments financed by public funds in order to improve the quality of spending. The processes have more or less come to an end depending on how old or how effective they have been. Within EU25 it is undoubtedly the UK that has had the most satisfactory return on experience with notably a significant development of allocated objectives and indicators to expenditure (reduction of their number, attempt to neutralise the perverse effects caused by the piloting of employment according to single indicators, reliability of results).

These processes are based on the notion that it is more effective both from an economic and from a strictly public finance point of view to provide rules to allocate means efficiently to public spending programmes rather than to proceed by "sectioning" the budget, since this does not depend on an analysis of spending. It is based on a reduction in budgetary fragmentation, with the identification of expenditure blocks that are greatly reduced in comparison with the old budget lines and which can be complemented with a periodical review of costs and advantages of the spending programmes.

In France, the organic law on finance bills (LOLF) voted in 2001 aims to transform the content of budgetary authorisation by re-organising it into "grand assignments" (34 in the draft finance law –PLF – 2006) which comprises the basis for public policies. Each assignment comprises programmes describing the end use of the expenditure and not its judicial nature; all of this is hinged on a performance measure that describes the objectives of the policies thereby financed, giving performance indicators and results targets; this measure is presented in annual performance projects that must face performance reports established after the execution of the budget.

But we have to admit however that the effects of such an approach (in France's case the law fully entered into application in preparation for the PLF 2006) will not be effective unless the lessons of the performance reports presenting the programmes' results are applied and if beyond that the definition of objectives clearly involves a ranking of expenditure. The "process-based" approach supposes therefore a political design that is stronger that the quantitative approach and which is at least as substantial as the structural approach which must incidentally comprise a consequence thereof.

This implies then major involvement on the part of Parliament in budgetary control and a change of mentality in terms of budget construction in order to accept structural changes in spending: the process-based approach must aim to bring forth a clearer consensus.

Moreover the OECD has revealed notably in the case of Sweden the conditions for a successful control of public finance in the framework of a budgetary results pilot, using the examination of healthcare spending as a base: it is important to limit the risks of a surplus offer by linking the results approach to a quantitative approach (as undertaken by Austria in its hospitals).

Finally the Canadian example of an overall review of programmes undertaken in 1994 shows that the success of the "process-based" approach implies calling into question a significant part of public spending; to achieve the objectives it had set itself in terms of budgetary balance and new policies the government examined around 36% of spending (excluding transfer expenditure to individuals and the service of the debt [16].

This programme review that achieved spectacular results in terms of spending control also resulted in a reduction of capital expenditure which was triple that of operational expenditure (in part via the transfer of patrimony, but also via more controversial choices in maintenance and renewal of patrimony).

Any process-based approach, although it enables the challenge of restricted expenditure, must therefore be complemented by regular efficiency appraisals of public spending.

To conclude: rediscovering the sense of public spending and making reversible choices

Whatever the level of public spending, it is more or less "endured": it is both the reflection of past decisions or commitments and a dynamic tool in economic policy.

The study of the situation in EU25 shows that there is no correlation between the burden of public spending and its "endured" dimension: several European states have demonstrated that it is possible to control the state of its public finance and yet maintain the liberty to intervene in the economy if necessary, a high degree of socialisation of their economy, a reflection of the choice of society.

The most attractive approach for success appears to be oriented towards the decision making process in public spending, that enables the identification of priority spending and those which it is appropriate to reduce, thereby providing public spending with renewed capabilities to act. Its success must undoubtedly find expression in the combination of quantitative and structural approaches to control spending, given that the objectives are allocated to public spending, beyond its reduction sometimes presented as inexorable, and that all players in public spending are involved, over a period of several years...