Abstract
This paper presents an overview of the different methodologies that can be used to measure the amount of available slack in an economy. A comparison of these alternatives reveals that the evolution of the Nairu is crucial for understanding recent cyclical developments in the French economy. A comprehensive reform of the personal income tax has been assessed. This paper analyzes the expansion of high-tech activities, and reviews a number of policy issues connected with the development of an information and communication technology-based new economy.
II. The French tax system – recent developments and key issues7
A. Introduction
24. France is characterized by one of the highest tax burdens in the industrialized countries, with a ratio of revenue to GDP some 20 percentage points higher than in Japan and the United States, and also above that of other euro area countries (Figure II.1). Even though differences in tax burdens partly reflect institutional arrangements (e.g., private versus public pensions and health care), France’s high tax burden is widely, and indeed officially, recognized to act as an impediment to sustained growth, notably by discouraging labor supply and investment. Prompted in part by these concerns, and the with the declared aim of ensuring sustainable inflation-free growth and of reducing unemployment traps, the government announced in late August 2000 a package of tax reductions, amounting to about F 120 billion (1¼ percent of GDP) to be implemented over 2001-2003. Combined with a number of tax cuts introduced in the 1999 and 2000 budgets, this package entails a significant reduction in a number of taxes and in the overall tax burden.
Government Revenue as a Percentage of GDP, 1999
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: OECD, Analytical Database.Government Revenue as a Percentage of GDP, 1999
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: OECD, Analytical Database.Government Revenue as a Percentage of GDP, 1999
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: OECD, Analytical Database.25. This chapter examines the extent to which the French tax system might hinder long-term economic growth and identifies priority areas where reform should concentrate in order to alleviate supply-side constraints to higher growth. After taking stock of recent reforms (Section B), the chapter provides an overview of the main characteristics of the tax system and a first broad assessment of the overall tax structure (Section C). The tax system is likely to influence long-term growth primarily through the incentives and/or distortions imposed on the behavior of workers and firms, and the chapter focuses on three main issues from this perspective. First, since reducing structural unemployment and boosting labor supply are crucial to increasing potential growth, the impact of the tax-benefit system on the performance of the labor market is examined in Section D. The specificities and weaknesses of the personal income tax (PIT), which is at the center of recent reforms, is then reviewed in Section E. Finally, the impact of corporate taxes on investment is examined in Section F. Section G provides some conclusions.
B. Overview of Recent Reforms
26. In the early to mid-1990s, tax policy in France was characterized by a mix of revenue-raising measures geared toward achieving the Maastricht deficit criterion (e.g., the 1995 and 1997 surcharges on corporate taxes, VAT rate increase) and reforms to reduce structural unemployment (e.g., targeted rebates on social security contributions). In recent years, the priority has shifted toward reducing the tax burden, first to sustain the economic recovery and then to correct the increase in the tax burden induced by the drive to EMU participation and a strong cyclical upswing. The budgets for 1999 and 2000 thus included a number of tax reductions amounting to 1.1 percent of GDP (Box II.1).
27. In addition, the government announced on August 31, 2000 a package of tax reductions and reforms, amounting to an overall F 120 billion (1¼ percent of GDP) to be implemented over 2001-2003. The main measures are:
A reduction in all the rates of the personal income tax: the lowest marginal rate will thus be reduced from 9.5 percent in 2000 to 7 percent by 2003, while the highest marginal rate will be cut from 54 percent to 52.5 percent by 2003. The overall cost of these measures will be about F 45 billion. In addition, the discount (“décote”) that applies to taxable income below a certain threshold will be modified to reduce the high marginal rates induced by the threshold.
The elimination over three years of the 1995 surcharge on the corporate income tax, lowering the normal statutory rate from 36.7 percent to 33.3 percent. In addition, the statutory rate is reduced in stages to 15 percent on the first F 250,000 of profits for small and medium-sized enterprises. These measures will be partly financed by changes in the rules for the taxation of dividends between subsidiaries, depreciation allowances and imputation, so that the overall cumulative cost will be F 20 billion.
The introduction of a graduated rebate on the contribution sociale généralisée (CSG) and contribution pour le remboursement de la dette sociale (CRDS)8 for wages up to 1.3 times the SMIC aimed at increasing in-work take-home pay for low-income workers already exempt from the income tax. This measure will cost about F 25 billion.
A reduction in excise taxes on domestic fuel oil to lessen the impact of increases in crude oil prices on household income. In addition, excises on gasoline will be adjusted to offset increases in VAT due to changes in oil prices, and the annual tax on private cars (vignette automobile) is to be eliminated.
Tax Reductions in 1999-2000
The most important tax-reduction measures introduced in 1999 and 2000 (amounting to 1.1 percent of GDP) were:
1999 Budget: F 41 billion (though F 25 billion was raised by a new ecotax and a surcharge on the corporate tax for large enterprises to finance the 35 hour work week), of which
a reduction of VAT on electricity and gas network access, appliances for handicapped persons, sorted waste treatment, and renovation of social rental housing (F 4.5 billion);
a reform of the taxe professionnelle (local business tax), by phasing out over 5 years the payroll component from the base (F 10.4 billion); and
an extension of the rebate on employer social security contributions for low-wage workers to 1.8 times the SMIC for enterprises implementing the 35-hour workweek (F 25 billion).
2000 Budget: F 40 billion, of which:
a reduction of VAT on home improvement and personalized care services (services de proximité) (F 20 billion);
a reduction of the real estate transaction tax (droits de mutation) (F 1 billion);
a broadening of the partial exemption of wages from the local business tax (F 2 billion); and
the elimination of the 1997 surcharge on corporate taxes (F 12 billion).
2000 Supplementary Budget: F 50 billion (F 40 billion of these cuts offset the carryover from 1999, higher projected growth in 2000, and larger-than-expected non-tax revenues), of which:
a reduction of the VAT rate from 20.6 percent to 19.6 percent as of April 2000 (F 18 billion in 2000 and F 31 billion for a full year);
a reduction of the taxe d’habitation, including the elimination of the regional tranche and new exemptions for low-income families (F 11 billion); and
a reduction in the two lower rates of the personal income tax, which also reduces the tax base by about 650,000 families (F 11 billion).
C. Overview of the Tax System
28. A breakdown of the tax burden into its main components shows that levies assigned to social security administrations, which include health care (CNAM), unemployment insurance (UNEDIC), pensions (CNAV) and family allowances (CNAF), amount to almost half of the total revenue collected (Table II.1). The remainder is accounted for mainly by central government taxes, as local government taxes are relatively small. In terms of tax instruments, indirect taxes (VAT, taxes on oil products, and other excises) represent a large share of revenue (almost 25 percent of the total), while the share of the personal income tax is relatively small. However, if the CSG is added to the personal income tax the proportion of taxes on personal income amounts to about 16 percent of total revenue, broadly in line with industrial country experience.
Structure of General Government Revenue, 1999
Excluding tax transfers from central to local government and Social security.
Mostly local indirect taxes and tax transfers from central government.
Specific taxes, transfers from BAPSA and duties on tobacco and alcohol.
Structure of General Government Revenue, 1999
| In percent of GDP | In percent of total | ||
|---|---|---|---|
| Central government | 18.7 | 40.9 | |
| Value added tax | 7.6 | 16.6 | |
| Excises on oil products (TIPP) | 1.8 | 3.9 | |
| Personal income tax | 3.4 | 7.4 | |
| Corporate income tax | 2.6 | 5.7 | |
| Other1 | 3.3 | 7.2 | |
| Local government | 5.5 | 12.0 | |
| Local business tax (Taxe professionnelle) | 1.5 | 3.3 | |
| Real estate tax (Taxe foncière) | 1.1 | 2.4 | |
| Housing tax (Taxe d’hetbitation) | 0.7 | 1.5 | |
| Other2 | 2.1 | 4.6 | |
| Social Security | 20.9 | 45.7 | |
| CSG | 4.0 | 8.8 | |
| Social contributions | 16.0 | 35.0 | |
| Other3 | 0.9 | 2.0 | |
| European Union | 0.6 | 1.3 | |
| Total | 45.7 | 100.0 | |
Excluding tax transfers from central to local government and Social security.
Mostly local indirect taxes and tax transfers from central government.
Specific taxes, transfers from BAPSA and duties on tobacco and alcohol.
Structure of General Government Revenue, 1999
| In percent of GDP | In percent of total | ||
|---|---|---|---|
| Central government | 18.7 | 40.9 | |
| Value added tax | 7.6 | 16.6 | |
| Excises on oil products (TIPP) | 1.8 | 3.9 | |
| Personal income tax | 3.4 | 7.4 | |
| Corporate income tax | 2.6 | 5.7 | |
| Other1 | 3.3 | 7.2 | |
| Local government | 5.5 | 12.0 | |
| Local business tax (Taxe professionnelle) | 1.5 | 3.3 | |
| Real estate tax (Taxe foncière) | 1.1 | 2.4 | |
| Housing tax (Taxe d’hetbitation) | 0.7 | 1.5 | |
| Other2 | 2.1 | 4.6 | |
| Social Security | 20.9 | 45.7 | |
| CSG | 4.0 | 8.8 | |
| Social contributions | 16.0 | 35.0 | |
| Other3 | 0.9 | 2.0 | |
| European Union | 0.6 | 1.3 | |
| Total | 45.7 | 100.0 | |
Excluding tax transfers from central to local government and Social security.
Mostly local indirect taxes and tax transfers from central government.
Specific taxes, transfers from BAPSA and duties on tobacco and alcohol.
29. As in a number of OECD countries, the revenue ratio has increased over the last two decades in France, reflecting a trend increase in social security contributions and taxes to finance rising transfers and subsidies (see IMF, 1996). In recent years however, there has been a marked decrease in the share of social security contributions (from 18 to 16 percent of GDP between 1997 and 1999), reflecting a shift from social contributions to the CSG (Figure II.2). Apart from this change, the tax structure has remained remarkably stable over the last decade. The main specificities of the French tax system appear to be a relatively small share of direct taxes on households and a high share of indirect taxes (Table II.2). Direct taxes on business also appear to be in the low range, although this may be partly offset by differences in social security contributions.
France Structure of General Government Revenue
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: Rapport économique, social et financier.France Structure of General Government Revenue
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: Rapport économique, social et financier.France Structure of General Government Revenue
Citation: IMF Staff Country Reports 2000, 148; 10.5089/9781451813517.002.A002
Source: Rapport économique, social et financier.Structure of Revenue in Selected OECD Countries, 1999
(In percent of total revenue)
Structure of Revenue in Selected OECD Countries, 1999
(In percent of total revenue)
| Social security | Indirect | Direct taxes | Direct taxes | Total | |
|---|---|---|---|---|---|
| contributions | taxes | on business | on households | ||
| France | 39.5 | 34.3 | 5.9 | 20.3 | 100.0 |
| Germany | 44.1 | 28.3 | 3.6 | 24.0 | 100.0 |
| United Kingdom | 19.9 | 36.9 | 9.9 | 33.3 | 100.0 |
| US | 23.6 | 25.7 | 9.3 | 41.4 | 100.0 |
| Japan | 39.4 | 31.7 | 10.1 | 18.9 | 100.0 |
Structure of Revenue in Selected OECD Countries, 1999
(In percent of total revenue)
| Social security | Indirect | Direct taxes | Direct taxes | Total | |
|---|---|---|---|---|---|
| contributions | taxes | on business | on households | ||
| France | 39.5 | 34.3 | 5.9 | 20.3 | 100.0 |
| Germany | 44.1 | 28.3 | 3.6 | 24.0 | 100.0 |
| United Kingdom | 19.9 | 36.9 | 9.9 | 33.3 | 100.0 |
| US | 23.6 | 25.7 | 9.3 | 41.4 | 100.0 |
| Japan | 39.4 | 31.7 | 10.1 | 18.9 | 100.0 |
30. Another approach to assessing the overall structure of the tax system consists in trying to compare it with an “optimal” tax structure derived from public finance theory. Although the theory does not provide many insights on the detailed desirable structure of taxation, one important result (Atkinson and Stiglitz, 1976) is that when an income tax is available, it is optimal that indirect taxes be neutral from a redistributive point of view and be used only to finance general public spending.
31. Bourguignon and Bureau (1999) note that the main taxes and expenditures in France can be classified into the redistributive function (ensured mainly by the combination of the personal income tax and the non-conditional benefits, or “generalized income tax”) and the all locative function (general expenditures financed by indirect taxes).9 They also observe that indirect taxes are broadly neutral in terms of redistribution, in the sense that the apparent rate of taxation varies very little with the level of revenue, and may therefore be considered as proportional taxes. They conclude that the overall structure of the tax system is not out of line with what the theory would recommend.10 However, the authors also underline the limits of such an overall assessment. In particular, this analysis focuses on the main aggregates and overlooks the specifics of the tax system that may lead to distortions within each of the main functions. In particular, they note that the degree of redistribution achieved by the “generalized income tax” is relatively limited compared to other countries, due to the narrow base of the personal income tax. Another limitation of this analysis is that the Atkinson-Stiglitz “optimality” theorem assumes that all markets are perfectly competitive, whereas in France the existence of the minimum wage and the policy of maintaining its purchasing power have to be taken into account when evaluating the tax system and designing tax reforms.
D. The Impact of the Tax-Benefit System on Labor Market Performance
32. The tax-benefit system may affect the functioning of the labor market in a number of ways: social security contributions and taxes create a tax wedge between employers’ labor costs and net after-tax wages available to employees, which affects labor demand and/or supply; the tax-benefit system influences the level of reservation wages and thus labor supply. In the 1990s, policies aimed at reducing structural unemployment focused on lowering labor costs for targeted categories of workers—mostly unskilled workers.11 These measures were based on the idea that the demand for low-skilled labor was constrained by a relatively high minimum wage and high social charges, possibly combined with skill-biased technological progress. In these conditions, and with an elasticity of labor demand generally considered to be higher for low-skilled than for high-skilled labor, shifting the tax burden away from low-skilled labor would induce a reduction in unemployment over the medium term.12
33. The annual cost of these measures is estimated at about F 40 billion (0.4 percent of GDP). Most ex-ante estimates evaluate the impact in terms of job creation at about 200,000 to 250,000 jobs over a five-year period. When taking into account the financing of the measure (e.g., through an increase in VAT or higher contributions on high-skilled labor), the estimate of the number of jobs created or maintained is reduced to 40,000 to 200,000 jobs, depending on the assumptions retained (CSERC, 1996). Since these measures were introduced in the first half of the 1990s, they are likely to have reached their maximum effect in terms of job creation by now, although the unusually high pace of employment growth in the current recovery could indicate that the impact of these measures has been deeper than anticipated. Very optimistic scenarios, based on a high elasticity of the demand for low-skilled labor and favorable supply-side effects, could thus lead to stronger estimates of the number of jobs created, up to 450,000 (before financing) in some studies. In this case, an appreciable reduction in structural unemployment might be underway.
34. Ex-post estimates of the impact of the measure are still very tentative but point to two significant developments. First, the long-term decline in the share of low-skilled workers in total employment (wage earners) observed in the 1980s was brought to a halt in the mid-1990s. Second, the proportion of jobs paid less than 1.3 times the SMIC picked up markedly in the late 1990s, which can be interpreted as the result of a larger number of jobs created in this wage range, or a limited diffusion of the discretionary increases in the minimum wage over the period. This increase in the proportion of jobs paid below 1.3 times the SMIC could also indicate that the rebates induced the development of “low-wage traps”, as the marginal cost for employers of increasing wages above 1.3 times the SMIC would be amplified by the loss of the rebate (Gubian, 1999).13
35. Although these approaches to reducing unemployment have been widely debated in Europe (see for example EU, 1994), other countries appear to have been reluctant to follow the French lead. In particular, Germany has limited this kind of measure to jobs paying less than DM 630 for less than 15 hours work, perhaps because of doubts about their cost-effectiveness. Indeed, estimates for France show that the cost per job created is relatively high, and some authors argue that the effectiveness of wage subsidies depends crucially on the pay-off to experience and training (Bell, Blundell and Van Reenen, 1999).
36. Although graduated rebates on social contributions have been the cornerstone of the French approach to reducing unemployment through tax reforms, a few other measures are worth mentioning, including the phasing out of the wage bill from the base of the local business tax, and the reduction in VAT rates on labor-intensive services. On the latter, Bourguignon and Bureau (1999) note that since social contribution rebates affect labor demand more directly than VAT cuts, they are likely to be more efficient, except possibly in a few very competitive and labor-intensive sectors where the price elasticity of demand is very high. They therefore argue for limiting targeted VAT cuts to activities that are at the margin of the informal economy or that could benefit from positive externalities.
37. Overall, the policies pursued in the 1990s appear to have been successful in increasing labor demand for low-skilled workers and thus possibly reducing structural unemployment. However, now that the economic recovery is well under way and that tensions are appearing on the labor market, the effects of the tax system on labor supply have acquired increased policy relevance. In this regard, of key importance is not only the level of taxes but also the combination of taxes and benefits for each level of income.
38. A useful measure of the extent of distortions imposed by the tax-benefit system is provided by effective rates of taxation, taking into account all relevant taxes and benefits. The marginal effective tax rate (METR) measures the share of any increase in a household’s gross wages that is taken away through increased taxes or reduced benefits.14 A number of recent studies have pointed out the existence of high METRs at the lower and higher ends of the income distribution (see also Box 5 of the Staff Report for the 2000 Article IV consultation, SM/00/217).
39. Bourguignon and Bureau (1999) estimate that the METR for a couple with two children (taking into account social contributions, the personal income tax, the CSG, the VAT, and transfers15) is higher than 100 percent for levels of income around the RMI (the revenu minimum d’insertion, the main basic income support scheme), due to the sharp withdrawal of some benefits, decreases to about 50 percent for wages around 2 to 4 times the SMIC, before increasing progressively to above 70 percent for wages above 15 times the SMIC. The authors note that while this U-shaped distribution of marginal rates may be seen as broadly in agreement with the theory of optimal taxation (Mirlees, 1971), it raises a number of problems. First, high marginal rates at the lower end may cause inactivity traps and, in a dynamic setting, prevent individuals from acquiring on-the-job training and experience that would increase their productivity. They note, for example, that a full-time job paying the SMIC increases monthly disposable income for a beneficiary of the RMI by only about F 1,700 (the monthly SMIC is about F 5,000). At the other end of the revenue distribution, they note that the scarce available evidence on the disincentive effects of high marginal rates is not conclusive, but that marginal rates above 75 percent are likely to be dissuasive (though emigration is likely to be influenced by differences in average rates of taxation across countries and not only by high marginal rates—see Section E below).
40. Laroque and Salanié (1999) estimate the proportion of households affected by the various levels of marginal rates, using data from the employment survey. They find that about 20 percent of the population face marginal rates higher than 90 percent, mostly at the lower end of the income distribution. Their simulations also show the diversity of marginal rates within a given range of income, reflecting the impact of family situations, composition of income, type of housing, etc. Finally, they estimate that less than half of the unemployed would gain more than F 3,000 when taking a full-time job paid at the SMIC, and about 40 percent of them would gain less than F 2,000. Going further in the analysis of labor supply, Laroque and Salanié (2000) break down non-employment between “voluntary” (people with a high reservation wage or unwilling to work for family, health, or other reasons), “classical” (people unable to find a job due to low productivity relative to minimum wage), and unexplained non-employment. Using participation equations estimated on data from the 1997 employment survey, they find that participation on the labor market largely depends on unexplained factors and varies significantly with the sex and marital situation. Overall, they estimate that 57 per cent of the non-employed are in the first category (“voluntary”), and that the participation of women on the labor market is more sensitive to financial incentives. All these estimates point to “inactivity traps” being quantitatively significant.
41. Calculations made by the Direction de la Prevision (Bourguignon, 1998, Annexe A) show more precisely that the effective marginal rate of taxation exhibits a number of peaks and irregularities that give rise to significant distortions. The marginal rate of taxation is thus 100 percent for a couple with two children with revenues equal to the RMI; it increases sharply as income rises above the RMI, due to the sharp reduction in housing aids that take into account income from work, but not from the RMI; then between the SMIC and 1.3 times the SMIC, the “ristourne degressive” induces an increase in the marginal rate; around twice the SMIC, the steep decline in housing subsidies and the personal income tax induce another peak; finally, each income bracket of the personal income tax results in a step increase in the marginal rate.
42. A number of measures have recently been taken to reduce or eliminate some of these peaks in the marginal rate. In particular, the reform of housing aids announced in the first half of 2000, during the last “Conference Famille” corrects the difference of treatment between work and non-work income and between the main subsidies (allocations logement and aide personalisée au logement) by creating a harmonized schedule based on overall resources. The reduction and reform of the housing tax also reduces the associated peak in marginal rates.
43. Although these measures correct the main irregularities, marginal tax rates remain high at the lower end. Various schemes may be considered to address this issue. A first natural option is to reduce the level of benefits, thereby increasing the difference between, for example, the RMI and the SMIC. Alternatively, a negative income tax, consisting of a flat benefit (independent of the level of income) combined with a proportional tax has been proposed. Such a mechanism results in a constant increase in the slope between gross work revenues and net disposable income. However, to ensure a minimum flat benefit of the same order of magnitude as the current RMI, either the marginal rate of the income tax or the cost of the program would have to be very high. Another possibility is the allocation compensatrice de revenu (ACR) which consists of a constant income support benefit combined with a flat rate tax with a rate lower than 100 percent applied to wages up to the SMIC, or 1.5 times the SMIC, and combined with the existing income tax for higher revenues. Bourguignon and Bureau (1999) suggest that such a system could be financed by a small increase in marginal rates for medium and high incomes or in the CSG. Such a system would have some similarities with the Earned Income Tax Credit in the United States (EITC) or the Working Family Tax Credit (WFTC) in the United Kingdom. In the EITC, a tax credit (which gives right to cash payment when it exceeds the tax liability) is granted to households with low incomes: the credit increases proportionally with earned income at first, then it becomes constant, and finally it declines to zero. An important feature of the EITC or ACR is that the benefit component is administered through the tax system. Although this may raise some administrative difficulties (e.g., the one-year lag associated with income tax filings, although it is noteworthy that few beneficiaries of the EITC in the United States have taken advantage of the possibility of receiving monthly payments), it also has the advantage of integrating various aspects of the redistributive system that are now divided between taxes and spending despite their close economic relationship.
44. In the event, the August 2000 tax reduction package eschewed any form of EITC, and opted rather for a graduated rebate of the CSG and CRDS paid on incomes up to 1.3 times the SMIC. Although all the specifics of the measure are not yet known (for example, how it will apply to part-time workers), it will not significantly lower the 100 percent marginal rate for people at the RMI.16 It will however reduce the marginal rates of taxation for people with earned income between the RMI and 1.3 times the SMIC, which should increase the incentives to work in this range. For example, SMIC earners will benefit from an increase in their net wage of about 10 percent. Although the recourse to the CSG may reduce the administrative costs of the measure, it also transforms a broad-based flat rate tax into a partly progressive tax, thus making the tax system more complex.
E. Main Weaknesses of the Personal Income Tax
45. Reform of the personal income tax (PIT) has long been on the agenda. Indeed, despite some important changes over the last decade, the PIT suffers from a number of well-identified weaknesses. Personal income taxation in France comprises two main taxes, the progressive personal income tax (IRPP), and the flat-rate CSG. Revenue collected by the personal income tax has declined from about 5 percent of GDP in the early 1980s to some 3½ percent today, reflecting increases in the number of exemptions and tax credits, which narrowed the tax base, as well as reductions in tax rates in 1993 and 1996. On the other hand, the rate of the CSG, introduced in 1991, was raised considerably in 1997–98 (to 7.5 percent) and revenue from this source now amounts to some 4 percent of GDP.
46. Although both these taxes can be considered direct taxes on income, there are important differences between them:
The CSG is earmarked to financing social security while the PIT finances general spending of the central government;
The CSG has a much broader tax base than the PIT: both are levied on labor income, pensions, and capital income, but the CSG applies to the first franc of revenue (after a 5 percent rebate for wages), whereas the PIT has a number of rebates and deductions (quotient familial, décote) that result in total exemption for about half of all households. In addition, some savings instruments (Plan d’Epargne Populaire (PEP), Plan d’Epargne Logement (PEL), life insurance) are exempt from the PIT but not from the CSG;
The CSG is a contemporaneous withholding tax for wages and pensions, and is individualized, while the PIT is collected with a one-year lag on a household basis; and,
Finally, since 1997, part of the CSG can be deducted from the PIT. This was justified by reasoning that the 1997 increase in the CSG replaced social security contributions for health care, which were themselves deductible from the PIT.
47. The main flaws of the PIT are a narrow base and, consequently, a low yield, high progressivity at the both the lower and higher ends (which blunts work incentives), and complexity.17 These three weaknesses are to some extent related. The accumulation of exemptions and rebates over the years may be seen as an attempt to offset the impact of high marginal rates. But they also have eroded the tax base while increasing the complexity of the system. Thus, a reduction in the tax rates would arguably be best combined with a streamlining of the web of exemptions and rebates. These include the exemption of some benefits, the 10 percent and 20 percent rebates on taxable income, which could be integrated in the tax schedule, and some tax credits. In addition, while the introduction of the CSG partly corrects for the small base and low yield of the income tax, the desirability of maintaining two complementary taxes on personal income remains questionable in the longer term. In addition, the tax deducibility of part of the CSG has no economic justification. Recent reforms of the PIT are somewhat disappointing in this regard, as they have focused mainly on reducing the tax rates, especially at the low end, without offsetting measures to reduce the number of exemptions and credits.18
48. The question of high marginal rates is more complex. At the lower end of the income distribution, the high marginal rate of taxation and associated inactivity traps stem from the benefit system rather than the PIT, given that low-income households are typically already exempt from the PIT (see section D for a discussion of these issues). For high incomes, the question is whether high marginal rates may lead high-skilled workers to emigrate, or might discourage the accumulation of human capital. However, the decision to emigrate should take into account relative average rates of taxation rather than marginal rates, and international comparisons show that these depend crucially on family situations. For example, when compared with the United Kingdom, the French PIT may be more costly for single high-wage people, but more advantageous for families with children. In addition, differences in the taxation of savings and wealth are also likely to play an important role in these decisions. Wealth is taxed in France when it is transmitted (droits d’ enregitrement) and on an annual basis through a wealth tax (impót de solidarité sur la fortune, ISF) and a property tax (impót fonder). The ISF applies to wealth (buildings, individual enterprises, financial assets, cars, planes, etc.) that is higher than F 4.7 million (net value), with progressive rates ranging from 0.55 to 1.8 percent. For taxpayers paying income tax in France, a ceiling limits the overall amount of income and wealth tax due to 85 percent of the previous year’s income.19 Although its effect is difficult to assess, it is generally recognized that the combination of PIT and ISF may contribute to emigration of wealth and taxpayers, and in particular, that the upper limit imposed on the tax rebate described above may have to be reconsidered in this light.
F. Corporate Income Tax
49. Corporate income tax (CIT) reforms have been relatively limited in France in recent years. The main measures have been the elimination of the two surcharges that had been imposed in 1995 and 1997, which will bring the standard corporate tax rate back to 33.3 percent by 2003, the ongoing phasing out of the wage bill from the base of the local business tax (taxe professionnelle), and recent targeted rate reductions for small- and medium-sized enterprises.20 Over the last 20 years there has been a widespread trend toward lower corporate tax rates in all industrial countries-although it was accompanied by a broadening of tax bases, leaving CIT revenues as a share of GDP broadly constant (Bond, Chennels, Devereux, Gammier, and Troup, 2000). A simple comparison of headline and typical corporate tax rates shows that France is in the middle of the range of industrial countries, with higher rates than in the United Kingdom, Denmark or the Netherlands, but lower than in Germany and Japan (Table II.3). It should however be noted that the French local business tax (taxe professionnelle) is not taken into account in these comparisons (whereas German local taxes are). Although the specifics of this tax make it difficult to compare with standard CIT, it probably contributes to increasing the cost of capital and the effective average tax rates above the levels reported in Table II.3.
Corporate Tax Rates
The German CIT reform will bring this rate from 40 to 25 percent in 2001.
Excluding surcharges.
Will be brought back to 33.3 percent by 2003 with the elimination of the surcharges.
Corporate Tax Rates
| Headline rates | |||||||
|---|---|---|---|---|---|---|---|
| United States | Japan | Germany | France | United Kingdom | Denmark | Netherlands | |
| 1979 | 46.0 | 40.0 | 56.0 | 50.0 | 52.0 | ||
| 1990 | 34.0 | 37.5 | 50.0 | 37.0 | 34.0 | 40.0 | 35.0 |
| 1999 | 35.0 | 30.0 | 40.01 | 33.32 | 30.0 | 32.0 | 35.0 |
| Typical tax rates | |||||||
| US | Japan | Germany | France | UK | Denmark | Netherlands | |
| 1979 | 49.6 | 52.6/40.0 | 62.2/45.0 | 50.0 | 52.0 | ||
| 1990 | 38.4 | 50.9 | 57.7/45.9 | 37.0/42.0 | 34.0 | 40.0 | 35.0 |
| 1999 | 39.3 | 40.9 | 51.6/42.8 | 40.03 | 30.0 | 32.0 | 35.0 |
The German CIT reform will bring this rate from 40 to 25 percent in 2001.
Excluding surcharges.
Will be brought back to 33.3 percent by 2003 with the elimination of the surcharges.
Corporate Tax Rates
| Headline rates | |||||||
|---|---|---|---|---|---|---|---|
| United States | Japan | Germany | France | United Kingdom | Denmark | Netherlands | |
| 1979 | 46.0 | 40.0 | 56.0 | 50.0 | 52.0 | ||
| 1990 | 34.0 | 37.5 | 50.0 | 37.0 | 34.0 | 40.0 | 35.0 |
| 1999 | 35.0 | 30.0 | 40.01 | 33.32 | 30.0 | 32.0 | 35.0 |
| Typical tax rates | |||||||
| US | Japan | Germany | France | UK | Denmark | Netherlands | |
| 1979 | 49.6 | 52.6/40.0 | 62.2/45.0 | 50.0 | 52.0 | ||
| 1990 | 38.4 | 50.9 | 57.7/45.9 | 37.0/42.0 | 34.0 | 40.0 | 35.0 |
| 1999 | 39.3 | 40.9 | 51.6/42.8 | 40.03 | 30.0 | 32.0 | 35.0 |
The German CIT reform will bring this rate from 40 to 25 percent in 2001.
Excluding surcharges.
Will be brought back to 33.3 percent by 2003 with the elimination of the surcharges.
In order to measure the impact of CIT on investment, it is necessary to look more precisely at the impact of taxation on the cost of capital (which measures the effects of CIT on “marginal” investment decisions) and/or at the effective average tax rates (which reflect the relative impact of CIT on the overall profitability of various investment projects). Bond and Chennels (2000) estimate the effect of CIT on the cost of capital and the effective marginal tax rates21 for seven countries (United States, Japan, Germany, France, United Kingdom, Denmark, and the Netherlands). This requires distinguishing between domestic and international investments (through a foreign subsidiary), and between various sources of financing (debt, equity, and retained earnings). Overall, their results show that France is generally in the middle to lower range in terms of both cost of capital and effective average tax rates, often close to U.S. levels, below Germany and Japan which exhibit relatively high costs and effective rates, and above the United Kingdom, Denmark and the Netherlands. In addition, debt financing appears relatively attractive in France compared to other countries.
50. At the same time, there remain important differences in the links between CIT and PIT across countries, although a movement away from full imputation (under which CIT paid can be deducted from PIT on dividends) seems to be under way in the European Union. Although this system presents the advantage of avoiding a double taxation of savings and reducing both average and marginal costs of equity-financed capital, difficulties in its implementation in an international setting appear to lie behind its discontinuation in the United Kingdom and Germany. France is thus one of the last countries in the European Union to retain full imputation (see the 2000 Article IV Selected Issues paper on tax reforms in Germany for a more detailed discussion of this issue (SM/00/229)).
G. Conclusion
51. To summarize, the French tax system exhibits the following specificities, which can serve as a basis for an agenda of further reform :
Targeted rebates on social security contributions have been successful in reducing constraints on labor demand resulting from high labor costs (SMIC and social charges) on low-skilled labor;
However, these measures are costly for the budget (especially after the extension to 1.8 times the SMIC) and may induce new distortions (e.g., low-wage traps);
The tax-benefit system still induces distortions and high marginal effective rates of taxation that are likely to limit labor supply, especially at the lower end of the revenue distribution.
The personal income tax system (PIT and CSG) is atypical and complex, with a combination of high marginal rates and numerous exemptions that reduce its transparency and its efficiency both in terms of yield and redistribution.
Corporate income taxes are, overall, not out of line with other OECD countries, but there is a need to remain attentive to developments on specific issues in an increasingly competitive international environment, such as the treatment of dividends, the impact of local taxes (taxe professionelle), and interaction between CIT and the PIT.
52. Several recent reforms have addressed some of these problems, in particular by smoothing the peaks in marginal effective tax rates related to the withdrawal of some benefits. More needs to be done, however, on both the benefits and the tax side to significantly reduce financial disincentives to work and therefore boost labor supply. In addition, a comprehensive reform of the PIT should tackle the problem of the small base and complexity of the PIT, and in the longer term, possibly consider moving toward a single withholding income tax. Finally, the increasing mobility of capital and high-skilled labor should lead to a continuous reassessment of the impact of taxation on investment and location decisions, including the taxation of savings and wealth. Finally, the analysis suggests that merely reducing the tax burden is unlikely to address these issues, which require deeper changes in the structure and specifics of some taxes and benefits.
References
Atkinson, A.B., and Joseph E. Stiglitz, 1976, “The design of tax structure: direct versus indirect taxation,” Journal of Public Economics, 6, pp.55 -75.
Bell, Brian, Richard Blundell and John Van Reenen, 1999, “Getting the unemployed back to work: the role of targeted wage subsidies,” The Institute for Fiscal Studies, Working paper No. W99/12.
Bond, Stephen and Lucy Chennells, 2000, “Corporate income taxes and investment: A comparative study,” Institute for Fiscal Studies, London.
Bond, Stephen, Lucy Chennells, Michael Devereux, Malcolm Gammie, and Edward Troup, 2000, “Corporate tax harmonization in Europe: a Guide to the Debate,” Institute for Fiscal Studies, London.
Bourguignon, Francois, 1998, “Fiscalité et redistribution,” Rapport du Conseil d’Analyse Economique, La Documentation Francaise, Paris.
Bourguignon, Francois and Dominique Bureau, 1999, “L’architecture des prélèvements en France : état des lieux et voies de réforme,” Rapport du Conseil d’Analyse Economique, La Documentation Francaise, Paris.
Cremer, David, 1999, “Fiscalité optimale et choix des instruments de la politique de taxation,” in “L’architecture des prélévements en France : état des lieux et voies de reforme,” Rapport du Conseil d’Analyse Economique, La Documentation Francaise, Paris.
CSERC, 1996, “L’allégement des charges sociales sur les bas salaries,” La Documentation Francaise, Paris.
European Commission, 1994, “Growth, Competitiveness, Employment: The Challenges and Ways forward into the 21st Century,” White paper, Luxembourg.
Gubian, Alain, 1999, “Six ans d’allègement de cotisations employeurs sur les bas salaires,” Bilan de la Politique de l’emploi 1998, Dossiers de la DARES, No. 3 -4.
IMF, 1996, World Economic Outlook, Chapter III, Fiscal challenges facing industrial countries, May 1996.
Laroque, Guy and Bernard Salanié, 1999, “Prélèvements et transferts sociaux: une analyse descriptive des incitations financeères au travail,” Economie et Statistiques, No. 328, 1999-8, pp. 3 -19.
Laroque, Guy and Bernard Salanié, 2000, “Une décomposition du non-emploi en France,” Economie et Statistiques, No. 331, 2000-1, pp.47 -66.
Malinvaud, Edmond, 1998, “Les cotisations sociales a la charge des employeurs : analyse économique,” Rapport du Conseil d’Analyse Economique, La Documentation Francaise, Paris.
Mirlees, John, 1971, “An Exploration in the Theory of Optimal Taxation,” Review of Economic Studies, Vol. 38, pp. 175 -208.
Sørensen, Peter Birch, 1997, “Public finance solutions,” Economic Policy, Vol. 2, Issue 25, pp. 221-264, October.
Prepared by Selma Mahfouz.
The CSG and CRDS are flat-rate taxes earmarked for the financing of Social Security.
Social security contributions can be seen as financing the insurance component of unemployment benefits, pensions, and health care.
See Cremer (1999), for a more detailed discussion of the optimal choice between direct and indirect taxation.
One of the key measures in this regard was the ristourne dégressive, consisting of tapered rebates to employers’ social contributions for low wages (below 1.3 SMIC); see also Box I.1 in Chapter I.
See Malinvaud (1998), on France, or Sorensen, (1997), for a discussion of the effects of these measures in a variety of theoretical models
The rebates were extended to 1.8 times the SMIC in the law on the 35 hour workweek, which should reduce (or displace) the problem of low-wage traps, but also weakens the targeting of the measure on low-skilled workers.
A marginal rate of 100 percent thus implies that the household’s total revenue does not increase at all, since all of the increase is paid in taxes or lost in benefits.
Transfers include the RMI (an income support scheme), family allowances, allowances for housing, and old age allowances.
As long as the calculation of the RMI différentiel is based on overall income excluding the CSG and CRDS (i.e., is based on net wages), the marginal rate will be unchanged for people who remain at the RMI after the measure, as their RMI would be reduced accordingly.
As noted above, the redistributive impact is limited by the narrow base and yield of the PIT.
An exception is the announced change in the “décote”(a discount that applies to taxable income below a certain threshold).
However, an upper limit has been imposed on this tax rebate for wealth above a given threshold.
In the context of the law on the 35-hour workweek, a new surcharge (contribution solidaire sur le bénéfices, CSB) was also imposed on larger firms.
These are forward-looking effective tax rates, calculated from the difference between the net present value of an investment project to shareholders with and without CIT.
