CHAPTER FIVE The Taxation of Returns from Personal Savings in France

Paul Masson
Published Date:
October 1995
  • ShareShare
Show Summary Details
Gary O’Callaghan

This chapter examines developments in France’s financial system since the mid-1980s, and their implications for the system of taxation of returns from saving.1 It will argue that the system is in need of a general review. The original approach to taxation of returns from saving has been overtaken by events, in particular by financial and capital liberalization and European financial market integration, and there has not been a systematic review of its goals and how they might be achieved. Instead, a number of specific taxation provisions have been introduced in a piecemeal fashion and without regard to changing financial circumstances, and have accumulated as a system that is overcomplicated and that frustrates its original (and subsequent) objectives.2 In addition, these provisions have severely reduced the revenues generated from the taxation of returns from saving without, at the same time, significantly increasing the overall incentive to save. The chapter outlines a general perspective that might be adopted in a review of the system and makes some specific policy recommendations.

A simple cross-country comparison reveals that the French system of taxation of returns from saving is subject to a greater range of exemptions and specific provisions than others.3 Many of these provisions existed prior to financial liberalization of internal markets in the mid-1980s and capital liberalization and European market integration after 1989, and are outdated. However, rather than overhaul the system, continued attempts to pursue a set of allocational goals through a system of differential tax treatment between financial instruments led to further exemptions in later years. The chapter will assess the combined impact of these exemptions in terms of achieving the overall goals of the system. In this sense, it will examine whether there has been an optimal response to a changing financial environment—has the system adapted appropriately to the additional constraints imposed by global financial developments? Has it adapted such that the existing structure of the system is consistent with its professed goals?

The summary review conducted here suggests three broad conclusions: First, the system as now constituted has some perverse effects that actually frustrate one of the policy goals—the goal of redistributing savings toward certain financial instruments. Second, attempts to reallocate toward specific instruments, by way of offering fiscal advantages, will be more distortional in a liberalized financial system and are less likely to be successful in promoting the second goal of policy, which is to foster an increase in aggregate savings. The third conclusion is that a uniform taxation system would not only be preferable in a liberalized financial setting, it would also reduce channels for tax avoidance and either lead to increased taxation revenues at existing rates or allow a general reduction of tax rates. The combination of the second and third conclusions would suggest that a uniform system would improve the trade-off between tax revenues and fiscal incentives aimed at increasing saving.4

These arguments, among others, are presented in more detail (in Section III) after a brief description of the development of the French system of taxation of returns from saving has been presented (in Section I) and the present system has been described (in Section II) and compared with others in the major industrial countries. Section IV offers some conclusions and specific policy recommendations.

I. The Evolution of the Current French System

Two objectives have been associated with the system of taxation of returns from saving in France: to increase the overall level of savings by offering tax advantages; and to use tax advantages as a means of reallocating savings towards specific instruments and uses (France (1993)). Prior to financial liberalization in the mid-1980s, this was attempted in the context of a banking system that was heavily regulated. Interest was (and is still) not permitted on checkable sight deposits, and most other savings instruments earned interest at regulated rates that were, for long periods, unchanged.

Until 1985, some 70 percent of the broadest (noncheckable and noncurrency) liquidity measure (L-M1) comprised savings instruments at regulated rates (see Chart 1). M2-M1, which comprised interest-bearing noncheckable sight deposits and noncheckable passbook savings accounts, made up almost 60 percent of L-M1 and all of the funds in M2-M1 were at regulated rates. In addition, virtually all of L-M3 was composed of contractual savings that were also subject to regulated interest rates. Only the term deposits in M3-M2 earned interest at rates that were flexible and non regulated.

Chart 1.France: Financial Aggregates

(In percent of L-M1)

Source: Banque de France, Bulletin Trimestriel (various issues).

Under the earlier system, with interest rates heavily regulated, many fiscal incentives were offered to specific savings instruments in order to allocate savings among alternative uses. This also meant that there was an incentive to allocate funds to specific financial institutions that offered these instruments. In 1984, less than one fifth of M2-M1 was subject to taxation while virtually all of L-M3 comprised instruments that were exempt from taxation. In 1985 taxation of income from financial capital reached an historic high at some 0.72 percent of GDP and 8.8 percent of central government direct taxation (France (1992)).

From 1985 onward, funds flowed increasingly into a range of newly created financial instruments and, in particular, certificates of deposit included in M3-M2, and treasury bonds and bills included in L-M3. By 1988, the proportion of M2-M1 in L-M1 had fallen below 45 percent, with M3-M2 having increased its proportion to 35 percent, and most of the funds that contributed to a doubling of the proportion of L-M3 (to 20 percent) were also at flexible interest rates. Clearly there had been a fundamental change in the French financial system, with more funds assigned to instruments at flexible interest rates, and this was compounded by increased European financial integration and the ability of capital to flow across borders following capital liberalization in 1989–90.

In 1989 the withholding tax rate on interest from savings deposits was reduced from 45 percent to 35 percent, and the withholding tax rate on income from most bonds and commercial paper was reduced from 25 to 15 percent. In addition, the taxation of life insurance benefits was virtually abolished. These reforms were enacted in preparation for capital liberalization and the integration of markets for financial services in Europe, and were prompted by an attempt to bring basic rates in line with the rest of Europe.5 Also, the marketing of foreign mutual funds that capitalize the financial income received as a means of delaying taxation was permitted in Europe from October 1, 1989. This was expected to lead to a displacement of financial intermediation outside of France and, in response, French financial institutions were permitted to sell such instruments. These added substantially to the proportion of savings in instruments at flexible interest rates, in particular money market mutual funds, because of relatively high short-term interest rates.

However, rather than overhaul the system in the light of these changed circumstances, the old practice of legislating fiscal preferences for different types of savings was extended through a series of stopgap measures. Furthermore, these have been adopted over the years without full consideration of their overall impact. A number of additional tax-exempt instruments were created in response to the increase in money market mutual funds, for example, in an attempt to lengthen the maturity of savings, and these are described in detail below. However, innovations aimed at securing a particular objective have often led to results that conflicted with earlier objectives, and have resulted in the introduction of successive stopgap measures to rectify the resulting distortions.

These developments also contributed substantially to an increase in the proportion of savings assigned to tax-free financial instruments. By the end of 1992, two thirds of the (newly defined) broad financial aggregate P2-M1 was free of taxation (Chart 2).6 This compares with 40 percent in 1988. In addition, tax revenues from earnings on financial capital had halved, to 0.32 percent of GDP (from 0.62 percent) and 4.1 percent of direct taxation (from 7.7 percent).

Chart 2.France: Financial Aggregates

(In trillions of francs)

Source: Banque de France, Bulletin Trimestriel (various issues).

II. Taxation Systems in France and Other Major Industrial Countries

There are a number of salient features of the French system of expenditure and taxation that stand out from other major industrial countries (see Table 1). First, the levels of government revenues and expenditure are among the highest, each being 8–10 percentage points of GDP above average. Second, the relative incidence of taxation on labor income is the highest among the major industrial countries while the incidence of taxation on other income, including returns from saving, is lowest. Further examination of the structure of taxation reveals two important features: (1) taxes raised on labor income include the highest proportion of employer contributions, and the second highest contribution rates (OECD (1992)); and (2) taxes raised via a general taxation system, which includes the taxation of other forms of income and returns from saving, are very low. The common element to these is that they both reflect a failure to extract taxes based on an individual’s overall ability to pay and put a commensurately higher burden on purely labor income.7 This contention is supported by Kopits ((1992), Chart 2).

Table 1.Major Industrial Countries: General Government Revenues and Expenditure in 1990
FranceOther Major Industrial Countries1West GermanyUnited KingdomUnited StatesCanadaItalyJapan
(In percent of GDP)
General government revenues48.838.940.539.333.343.242.734.7
General government expenditure50.241.942.340.635.747.353.631.8
Social protection expenditure226.018.924.920.314.818.623.411.6
Memorandum items(In percent of total)
1.Structure of taxation
Labor income46.126.736.817.529.514.233.229.2
Employer contributions27.315.618.810.016.69.723.615.2
Other personal income11.830.927.428.435.840.826.326.8
Corporate income5.410.44.711.07.37.510.221.5
(In percent of GDP)
2.Present value of accrued pension rights—net liabilities21615215715689121242145
Sources: IMF, World Economic Outlook: Interim Assessment, January 1993; OECD, Economic Survey France, 1992.

Nonweighted average of major industrial countries excluding France.

Source: OECD (1992); for EU, data for 1988.

Sources: IMF, World Economic Outlook: Interim Assessment, January 1993; OECD, Economic Survey France, 1992.

Nonweighted average of major industrial countries excluding France.

Source: OECD (1992); for EU, data for 1988.

The other notable feature of Table 1 is the extent of unfunded liabilities to workers in the form of accrued pension rights in France. These are second highest (to Italy) among the major industrial economies and imply an alarming future burden on a taxation system that already places one of the highest burdens on an economy among the major industrial countries. Given the present structure of the tax system, with its relatively heavy reliance on taxes on labor income, this implies a growing taxation wedge on labor unless reforms to the system are coupled with a movement toward a more generalized, and broad based, taxation system.

The Structure of Taxation in France and the Major Industrial Countries

Aside from specific exemptions, the taxation of returns from saving is first affected by the overall structure of the taxation system. In this regard, France does not at first appear to be exceptional (Table 2). Dividend, interest, and capital gains taxes are included with other sources of income in the general taxation system, but this is subject to the exception that residents can opt for a final withholding tax on interest income. This also applies to the Italian system, while Japan completely excludes interest income from the general taxation system. In France, savers are allowed to choose the lower of the final withholding tax on interest or the marginal liability from the general taxation system, and this is an element of preferential treatment for interest income. Dividend income is allowed a 50 percent credit for corporate taxes paid, unlike in other countries, but is thereafter subjected to the marginal income tax rate. Capital gains are disadvantaged by being taxed at the marginal income tax rate and being allowed no deductions.

Table 2.Major Industrial Countries: Taxation Systems, 19911
FranceUnited KingdomGermanyItalyUnited StatesCanadaJapan
Included with other income
Capital gainsYes2YesShort-term3Yes4Yes75Yes
Pension income90Yes60YesYes17No
Dividends550NoDM 6006NoNoNoLimited
InterestF 8,000NoDM 6006NoNoNo
Capital gainsNoVariousDM 1,000NoNoC$ 10,0007Y 500,0008
Withholding taxes on interest
Other deductions/credits
Pension payments
Government social securityYesNoLimitedLimitedNo17100
Private pension schemeYes15-4010NoLimitedNo17No
Life insurance25NoLimitedL 2.5mNoNoYes
Mortgage interest25Yes11Rental onlyL 7.0mYesNoNo
Sources: International Bureau of Fiscal Documentation (1992); and Price Waterhouse (1992).

Unless otherwise specified, numerical values are in percent of total relevant income.

Substantial capital gains (exceeding F 316,900) are taxable at a fixed rate of 16 percent.

Long-term gains on significant interests (25%) may be taxed.

Capital gains on shares are taxed at a flat rate of 25 percent.

In addition to an allowance for corporate taxes paid.

Total deduction is DM 600 for both dividends and interest.

Lifetime exemption.

Gain on long-term assets (5 years) receive 50 percent deduction. Some gains taxed at flat rate of 25 percent.


Percent of total income.

On loans up to £30,000.

Signifies combined deduction.

Sources: International Bureau of Fiscal Documentation (1992); and Price Waterhouse (1992).

Unless otherwise specified, numerical values are in percent of total relevant income.

Substantial capital gains (exceeding F 316,900) are taxable at a fixed rate of 16 percent.

Long-term gains on significant interests (25%) may be taxed.

Capital gains on shares are taxed at a flat rate of 25 percent.

In addition to an allowance for corporate taxes paid.

Total deduction is DM 600 for both dividends and interest.

Lifetime exemption.

Gain on long-term assets (5 years) receive 50 percent deduction. Some gains taxed at flat rate of 25 percent.


Percent of total income.

On loans up to £30,000.

Signifies combined deduction.

The structure of taxation in France in 1991 is outlined in Table 3. The general structure of the system has not changed since then, although some minor changes will be noted in the description that follows. Income tax (impôt sur le revenu) applies to income from employment, income from dividends and interest in excess of F8,000 per year (unless a final withholding tax is opted for, see below), and pension income and capital gains on the sale of assets for more than F 20,000 (unless a “substantial interest” is held and a flat rate is applied, see below). Social security payments are deductible from taxable employment income, and supplementary contributions to other pension plans are deductible subject to certain limits. While the social surcharges (Table 3, line 24) have not heretofore been deductible, an increase in the contribution sociale généralisée (CSG) in July 1993, to 2.4 percent from 1.1 percent, was made deductible.

Table 3.France: Personal Taxation Calculation, 1991
1.Employment income
2.Less: social security payments
4.Dividends and qualifying interest income
5.Less: exemptionFirst F 8,000
7.Capital gains on a nonsubstantial interest in shares or bonds1
8.Pension income
9.Less: exemption10 percent of (8)
11.Total(3 + 6 + 7 + 10)
12.Income tax (impôt sur le revenu)Calculated on (11)
13.Less: deductions:(14 + 15)
14.Mortgage interest25 percent of interest
15.Life insurance credit25 percent of premium up to F 1,000
17.Capital gains tax on a substantial interest in shares and bonds116 percent
18.Subtotal(16 + 17)
19.Less: minoration0-11 percent of (18) depending on level
20.Less: dividend tax credit (avoir fiscal)50 percent of dividends
21.Less: investment credit25 percent of investment up to F 2,500
23.Final withholding tax (prélèvement libératoire)15 to 50 percent of interest depending on debt instrument
24.Social surcharges:(25 + 26 + 27)
25.Contribution sociale généralisée1.1 percent on all income and capital gains
26.Complementary contribution1 percent on all income
27.Social tax1 percent on all income and capital gains
28.Total(22 + 23 + 24)
Source: International Bureau of Fiscal Documentation (1992); and Price Waterhouse (1992).

A substantial interest exists when transactions exceed F 316,900 per year or the taxpayer holds 25 percent or more of the shares.

Source: International Bureau of Fiscal Documentation (1992); and Price Waterhouse (1992).

A substantial interest exists when transactions exceed F 316,900 per year or the taxpayer holds 25 percent or more of the shares.

The top marginal income tax rate is 57 percent. However, this will not apply to most interest income because the taxpayer may elect to be taxed at a final flat rate (prélèvement libératoire, Table 3, line 23) that depends on the instrument involved. The rates range from 15 percent on bonds or other negotiable instruments to 35 percent on interest from savings deposits and 35–50 percent on interest from short-term claims on certain institutions (bons du trésor or bons de caisse). Income from foreign securities is generally only taxable at graduated rates. Dividend income receives a tax credit of 50 percent of the income received (the avoir fiscal, Table 3, line 20) in order fully to compensate for corporate taxes paid by the company (at the corporate tax rate of 34 percent).

Capital gains on securities representing a substantial interest (when aggregate transactions in securities exceed F 316,900 a year or the taxpayer holds 25 percent or more of the shares) are taxed separately from income and subject to a 16 percent flat rate (Table 3, line 17). Capital gains arising from the sale of real property are taxable at a rate that declines with the length of time for which the asset was held. The sale of a principal residence is tax exempt. There is also a 0.6 percent habitation tax on income, which is withheld to benefit local authorities, and a wealth tax of 0.5-1.5 percent is imposed on wealth exceeding F4,390,000.

Exemptions from Taxation of Returns from Saving

Exemptions from taxation of returns from saving are far more extensive in France than in other major industrial countries. Where there are exemptions in other systems, these generally apply to government savings instruments where the funds are at the disposal of the government, rather than to instruments in private institutions. The brief list of exemptions for the other major industrial countries is presented in Appendix I and should be compared with the following description of the extensive system of exemptions in France.

In addition to the general exemption on interest and dividend income (of F8,000) mentioned above, and the avoir fiscal on dividend income, there is a range of deposit interest income that is completely tax exempt (including from the social surcharges, line 24 of Table 3). The principal instruments are listed in Table 4, where the instrument is defined, and the maximum interest income or deposit that is free of taxation is noted.8 The relative magnitudes of the funds invested in the various instruments as at the end of 1992 are also supplied and some additional provisions are noted.

Table 4.France: Tax-Free Savings Instruments and Financial Aggregates, 1992
Monetary AggregateTax-Free InstrumentInterest RateCeiling (F ,000)Holding PeriodBalance (F bn.)Percent of AggregateGrowth Rate (in percent)1Remarks
Livrets A et bleus4.5100.0738.461.4-3.9Ordinary deposit accounts have similar rates.
Livrets CODEVI4.515. to benefit PMEs.
Comptes d’épargnelogement4.0150.01.5125.010.44.2After 1½ years, gain right to low-interest mortgage. Some interest paid by state.
Livrets d’épargne populaire5.5100.075.06.2Only low-income persons.
OPCVM (money market funds)325.81,482.357.214.3Ceiling halved from January 1, 1993.
P1 (Housing savings plans):1,044.2100.013.9
Livrets d’épargne entreprise (LEE)
Plan d’épargne- logement (PEL)6.0400.04.0482.546.24.3After four years, gain right to low-interest mortgage. Some interest paid by state.
Plan d’épargne populaire (PEP)600.08.0274.226.338.3To encourage long-term savings but can borrow against.
Insurance companies57.25.535.5
P2 (Bonds)2,138.0100.012.0
Life insurance reserves918.042.919.3Tax deduction of 25% of annuity up to F 4000.
OPCVM (bonds)325.8545.025.56.5
P3 (Shares)11,115.0100.01.6
OPCVM (shares)325.8195.01.8-2.0
Plan d’épargne en actions (PEA)600.06.0Introduced at end-1992.
Memorandum items
Percent of aggregates exempt from taxation2
Source: Banque de France, Annual Report 1992.

Nominal growth in 1992.

Assuming 80 percent of life insurance tax free.

Source: Banque de France, Annual Report 1992.

Nominal growth in 1992.

Assuming 80 percent of life insurance tax free.

The livrets A et bleus are designed to reallocate funds, via a government credit agency, to some socially desirable function. They each paid an interest rate of 4.50 percent as at the end of 1991. Comptes and plans d’épargnelogement are intended to encourage saving for housing and offer rates ranging from 2.75 percent to 6.00 percent.

CODEVI (comptes pour le développement industriel) accounts were created with the objective of providing financial and tax incentives for small and medium-sized firms (petites et moyennes entreprises, or PMEs). They are special tax-exempt accounts, which pay a fixed interest rate (4.50 percent as at the end of 1991), and whose funds are reserved for loans to PMEs. In August 1991 the ceiling on deposits was raised to F 15,000 (from F 10,000), and the rate of interest charged to PMEs was lowered to 8.75 percent (from 9.25 percent).

Mutual funds or OPCVM (organismes de placements collectifs en valeurs mobilières), comprising SICAV (sociétés d’investissement à capital variable) and FCPs (fonds communs de placement), have been established and exonerated from taxes on income earned since January 1989. Tax-free withdrawals are limited to F 325,800 per year (as at the end of 1992). High money market rates have led to a substantial increase in money market mutual funds, which reached F 1,300 billion in April 1993 (from about half that level at the beginning of 1990), and the tax-free limit (on withdrawals from money market funds only) was halved as of January 1, 1993.9 Withdrawals beyond the ceiling are taxable at the 15 percent flat rate (on negotiable securities) plus the three social taxes (including the CSG).

The substantial increase in money market mutual funds led to a concern that there had been a substitution of short-term for long-term savings. In response, a tax-free deposit at savings institutions (the plan d’éepargne populaire, or PEP) was introduced in 1990 to encourage long-term savings, but this meant that dividend income was then at a disadvantage. To rectify this, PEA (plan d’épargne en actions) mutual funds were introduced with similar incentives for share investment in 1992.

PEP accounts were introduced on January 1, 1990 (to replace an old age saving scheme) and experienced substantial growth in 1992. They are offered by (approved) banks and insurance companies only, and the funds cannot be withdrawn for eight years (an account can only be open for ten years). They were designed to promote long-term saving, earning an annual interest rate of 5.50 percent as at the end of 1991 (subsequently raised to 9 percent), and the maximum deposit is F 600,000 (or double for a married couple).

PEA mutual funds were introduced in October 1992, and were designed to encourage savings in shares. They are limited to F 600,000 per individual (or F 1.2 million per household) and are held with (approved) institutions. They are exempt from taxation on dividends and capital gains if held for six years and cannot be held for more than eight years. They were intended to “level the playing field” because the PEP had changed the incentive structure away from investment in shares.

The combination of the measures outlined above allows investors in France to avoid virtually any taxation on returns from saving. However, as is demonstrated in Section III, the system is unlikely to be successful in achieving its goals, and a general review of those goals and the manner in which they might be achieved is warranted.

III. A General Review of the System of Taxation of Returns from Saving

A few general points regarding the overall structure of the system, and the effect of changing financial circumstance, are warranted before the effect of exemptions for specific instruments is examined.

Overall Taxation of Returns from Saving

The application of a final withholding tax on resident interest income is also found in three other European countries—Belgium, Greece, and Portugal (Gardner (1992), p. 60). There are two aspects to the process of capital liberalization and financial market integration in Europe that have severely constrained individual countries in levying taxes on interest income of resident individuals, and prompted the introduction of final withholding taxes. First, the ability of funds to flow across borders without detection has been eased, and it is more difficult to apply the residence principle of taxation whereby taxes are levied by the country of residence on an individual’s global income. In these circumstances, an investor may only have to pay the foreign withholding tax on interest earned abroad and, unless there is a final withholding tax in the country of residence, resident investors will have to pay higher marginal income taxes at home. Unless there is an equivalent (or lower) withholding tax at home there will be an incentive for residents to invest abroad. The second problem is that, in the absence of harmonization of withholding tax rates on returns from saving across different countries, there may be competitive reductions of withholding tax rates as countries attempt to attract foreign funds. This would reduce the tax take for all countries (see Gardner (1992)).

France has responded by lowering withholding taxes, but still has one of the higher withholding tax rates on interest income in Europe. However, the increasing reliance on tax exemptions may also have been a response to the fact that withholding taxes in some countries (most notably Luxembourg) are set at zero. The efficacy of specific exemptions is examined below but, more broadly, agreement between EU countries on a minimum amount of withholding of tax or on the reporting by financial institutions would be desirable.10

With such considerations pertaining to interest income, and because capital gains and dividend income are taxed at marginal tax rates, interest income receives a significant tax advantage over other forms of revenue from savings. Retained corporate earnings, which should ultimately produce a capital gain for individuals when shares are sold at a higher price, are effectively taxed at the marginal rate of taxation (unless a substantial interest is held) and this is a disincentive to investing in shares.

Effects of Exemptions

Most studies of the taxation of returns from saving have been conducted at an aggregate level where an optimum uniform rate of taxation can, at least in theory, be determined. Taxing income from savings will generate revenue but will also reduce the availability of savings at any given interest rate and (by shifting a savings function inward) result in a reduction in capital accumulation and productive capacity. The optimum level of taxation will be determined in the context of this trade-off between tax revenue and productive capacity. It will also be influenced by similar decisions regarding the taxation of other bases (income, consumption, and wages) that cause other distortions. Auerbach and Kotlikoff (1987), for example, examine such questions in a dynamic setting where overall welfare is maximized by limiting the distortions to economic growth that are caused by the combined rates of taxation.

There is a second (interrelated) level at which decisions regarding taxation of income from savings must be made, however, and this is the degree to which preferential tax treatment is given to specific sources of returns from saving. This is particularly applicable to France, given its wide range of tax exemptions, and these have become increasingly distortional over time as the financial circumstances in which they were introduced have changed. A simple analytical model is presented in Appendix II and can be used to support four basic points relating to the adaptation of the French system of taxation of returns from saving to financial and capital liberalization.

First, the pressure of competition from foreign financial instruments (including SICAVs) bearing flexible interest rates has prompted the introduction of domestic competitors. These have been afforded tax exemptions similar to those of the pre-existing regulated instruments. Liberalized financial instruments bear higher interest rates, but have lower nonpecuniary advantages such as acquired rights to mortgage financing or simple convenience. However, because the same system of tax advantages is applied to both, and because this alleviates taxes on interest income, it will actually serve to increase the relative incentive to invest in the (higher-interest-bearing) liberalized instruments. This obviously frustrates the original intention of offering tax exemptions to regulated instruments.

Second, to the extent that a system of tax advantages for regulated instruments still exists, but in a liberalized financial setting, this causes larger distortions to preferences than existed in the old system where tax advantages were assigned solely to certain regulated instruments. This is because the range of instruments available has increased. Also, because the new instruments (certificates of deposit, for example) generally bear higher interest rates, there will be increased incentives to incur transactions costs and switch between instruments in order to take advantage of tax breaks, and then borrow against the regulated (and tax-exempt) instrument. It is possible, for example, to use a PEP as collateral against a loan. More generally, a system of preferential tax incentives is not compatible with financial liberalization that allocates funds on the basis of market incentives.

Third, to the extent that one of the goals of policy is to encourage increased savings by providing tax incentives, a system of specific exemptions will increase the incentive to incur transactions costs (as outlined above) and these will commensurately reduce the overall increase in the return to saving provided by the tax exemption. This results in a worsening of the trade-off between tax revenues and the levels of saving and capital accumulation. More significantly, perhaps, it can cause substantial economic losses for individuals by increasing the economic costs associated with acquiring the preferred set of instruments.

Fourth, the application of thresholds to fiscal exemptions can be self-defeating. Exemptions are secured by allocating funds to different financial instruments, up to the extent of the ceiling established, and this means that, while there is an incentive to allocate funds to specific instruments, there is also an incentive to allocate to others (with exemptions) when the ceiling is reached. Thus exemptions that aim to favor one instrument may frustrate the incentives to invest in another.

The original set of tax exemptions was introduced in an attempt to increase the effective return on savings available and reallocate savings toward some specific regulated instruments. This was successful in a regulated environment even though it did cause financial distortions. However, the successive introduction of financial and capital liberalization serves to increase the distortions caused by tax incentives and ultimately to frustrate the original policy goals. This is because a third goal of policy has been implicitly added to the original two, which is to limit the extent of intermediation occurring abroad, and the third is not compatible with reallocation toward a specific regulated domestic asset. Furthermore, financial liberalization involves the implicit expression of a fourth goal of policy, which is to allow capital markets to allocate funds freely and take advantage of market efficiency. All four goals cannot be simultaneously retained.

The lesson from this is that, because of the possibility of funds flowing to foreign (and effectively tax-free) financial instruments, the goal of reallocating toward specific domestic instruments must be relinquished. Henceforth, the maximum rate of taxation must be dictated by what is possible in the context of liberalized capital markets and, thereafter, the same rate of taxation should be applied to other instruments. If reallocation is attempted through lower tax rates on specific instruments, this will cause distortions that are far larger than in a regulated financial environment. The effectiveness of attempts to introduce specific incentives is diminished in a liberalized setting, which is a strong argument for a uniform system of taxation that will allow the relative benefits of instruments to attract investors. Overall, tax advantages will aggravate allocational distortions and are less likely to succeed in a liberalized financial environment.

The extent of returns from saving relief is accentuated by the range of instruments that are exempted. This allows savers to reallocate savings instruments in response to changing market and interest rate conditions. This has been a principal feature of the French system in recent years with large-scale fund reallocation in response to changing market conditions. There are many other fiscal incentives toward savings involving exemptions for life insurance contributions, pension contributions, mortgage interest relief, and flat rate taxes on capital gains on a substantial interest. These incentives generally take the form of exemptions from direct taxation and provide further avenues to protect income from savings from any real taxation. There are also subsidized interest payments on certain instruments (see Table 4).

There are a number of other distortions caused by a system of tax exemptions, which detract from an efficient allocation of savings and increase transactions costs.

A number of incentives (e.g., PEP and PEA) are offered on instruments that lock in funds for a particular time period. These are designed to increase the term for which savings are undertaken. This intention is, of itself, questionable on the basis that individuals may have a preference to maintain a flexible portfolio in case financial circumstances change. In reality, however, it is possible to borrow using these funds as collateral, and the lock-in effect only serves to increase transactions costs. This is a prime example of the type of problem referred to above.

Financial intermediaries are faced with difficult planning decisions because many of the specific instruments are offered only by specific financial institutions, and there are large flows of funds between instruments in response to changes in fiscal regulations and market conditions. For example, banks cannot offer livrets, and these are likely to become more popular as short-term flexible interest rates decline. This frustrates attempts at financial planning where large-scale balance sheet changes can occur in response to fiscal incentives.

IV. Conclusion and Recommendations

The overall conclusion from the foregoing is that the system of taxation of returns from saving in France is overcomplicated and produces less revenue than in other major industrial countries. At the same time, many of the goals of the system have been frustrated by financial and capital liberalization. The present system has been overtaken by financial developments but has not been subjected to review. Instead, a plethora of exemptions has successively reduced tax revenue from savings without necessarily achieving its assigned objective.

A complete review of the system should consider the following points: First, the goals of the system of taxation of returns from saving should be clearly expressed and ranked with reference to several considerations. If the system aims at securing a high level of revenues, this should be exercised in the most efficient manner, and would involve minimizing the set of potential loopholes when exemptions are employed to avoid taxes. At the same time, if the system aims to minimize the disincentives to saving, it should avoid introducing unnecessary transactions costs that are caused by the same exemptions. The elimination of exemptions would also enhance the efficient allocation of savings to their most desirable ends by allowing the liberalized financial system to operate. None of these intentions is consistent with attempts to allocate funds toward specific financial instruments, and this latter goal of policy is only achieved at a high cost in terms of sacrificing the others.

Second, in a liberalized financial system the practice of granting a range of specific tax exemptions can be counterproductive in terms of reallocating savings. Exemptions should be considered in a general context, where their full effects are taken into account, rather than in terms of the effects they would have in isolation. This would lead to a more general appreciation of their shortcomings.

Third, the most overriding argument for eliminating a system of specific tax incentives is that they can significantly increase transactions costs. These costs are likely to have increased with financial liberalization and detract significantly, and unnecessarily, from the benefits of market allocation.

Fourth, to the extent that an incentive is considered desirable for small savers, the universal tax exemption (of F 8,000) would serve as a more effective incentive. A reduction in transactions costs would increase the return from savings and this exemption could, if considered desirable, be extended.

Fifth, and more generally, a more uniform tax code should be used to reduce the disadvantage accorded to dividend income and retained earnings. This would be preferable to the introduction of schemes such as the PEA to encourage investment in shares.

Appendix I: Exemptions to Taxation of Returns from Saving in the Major Industrial Countries

Canada: Contributions to a number of registered savings plans are deductible for tax purposes.

Germany: In general, interest income is not exempt although there are some exemptions granted for interest deriving from certain favored sources, for example, interest from qualifying life assurances, and Government-inscribed debt. In general, an allowance of DM 600 is granted to individuals (DM 1,200 for spouses filing a joint return).

Italy: An exemption from taxes on income is granted with respect to interest on Treasury bonds, Post Office bonds, and some other public debt instruments. In addition, exemptions apply to some specialized public bonds issued under laws aimed at encouraging investment in certain industries or regions.

Japan: Interest received on certain government bond issues is exempt.

United Kingdom: Each individual is exempt from income tax on the first £70 of interest received on deposits at the National Savings Bank and on all interest received on limited holdings of National Savings certificates. With a personal equity plan (PEP) an individual may invest up to £6,000 per year in an investment fund PEP and £3,000 in a single company PEP, both investments being exempt from income tax and capital gains in the hands of the PEP investor.

United States: Interest on federal obligations of the United States has been subject to income tax since 1941 (with one minor exception). Interest on debt instruments issued by the states is exempt from federal taxes if they are issued for government activities, certain environmental activities, or for mortgage subsidies. Tax shelters can be used to offset taxes on passive investment income only.

Appendix II: An Analysis of Exemptions from the Taxation of Returns from Saving

This appendix will attempt to demonstrate why a uniform tax rate on returns from saving is preferable to a system of specific provisions and exemptions. It does this by developing simple measures of both the incentives and distortions generated by specific provisions in the same analytical framework. In general, attempts to reallocate savings toward specific financial instruments will be less effective in terms of increasing the return on savings, and will cause greater distortions, than a generalized tax exemption aimed at increasing the rate of return on savings. This result is likely to be strengthened with financial liberalization. Furthermore, attempts to reallocate toward a specific regulated financial instrument may be counterproductive.

Four points regarding the effectiveness of, and distortions resulting from, a system of specific provisions are demonstrated:

(1) a tax exemption introduced to benefit a particular financial instrument, even in a regulated financial environment, may not be fully reflected in an increased after-tax return on savings. The incentive effects will be offset to the extent that transactions costs are incurred in capturing the tax advantage but then reasserting individual preferences for an alternative instrument;

(2) a system of individual thresholds on tax exemptions, instead of a generalized exemption on all instruments, will increase the likelihood that these transactions costs are incurred and will therefore have a smaller impact on the rate of return on savings;

(3) the incentive to incur transactions costs, and therefore the level of transactions costs affecting the system, increases with the extent of financial liberalization; and

(4) the introduction of a system of generalized exemptions in a liberalized financial system can lead to results that are contrary to their original intentions.

A Regulated Financial Market—Instruments Bearing the Same Rate of Financial Return

Consider a setting in which there are two financial instruments (A and B) available. Each receives the same rate of (fixed) financial return (i*) but has, in addition, a different set of nonpecuniary advantages to the representative agent (respectively, σa and σb for each instrument). Assume, for simplicity, that the nonpecuniary advantages are constant (i.e. that they do not diminish with an increase in the holdings of a particular instrument). Then the representative agent will hold all of one instrument (there will be a corner solution), depending on the total (pecuniary and nonpecuniary) total returns (ra or rb), where

ra = i* + σa,

rb = i* + σb,


ra,b = ra - rb = σa - σb

is the relative return to instrument A.

Assume also that the overall level of savings will depend on the return received on the chosen instrument

Returns from saving are initially taxed at a rate τ but a tax exemption is offered to instrument B both to increase the overall level of savings and reallocate savings toward B. Therefore, the total after-tax rates of return are

ra = i*(1-τ) + σa

rb = i* + σb,

and the relative return on B becomes

rb,a = i*τ - (σa - σb).

Chart 3 (where the diagonal lines are at 45°) describes the options available to the investor under the assumption that σa - σb > 0. Initially, funds are allocated at point A. Then, with a tax exemption available on B, funds are switched to B. However, there is another alternative available, which is to take advantage of the tax exemption on B, and then borrow funds against this deposit and redeposit in A (to take advantage of higher nonpecuniary gains). If borrowing and lending rates are the same, and there are no intermediation or transactions costs, this would allow the investor to move to point A′. There will, however, be costs incurred on such a transaction (ϕ) and these will limit the investor’s ability to move to point A′. The higher the costs, the more the investor’s maximum possible return (at some point A″ to the left of point A′) is reduced. The maximum return is reduced to i* + σa - ϕ. If ϕ exceeds σa - σb (which is the return at point A′ less the return at point B) no switching for tax purposes will occur and the investor retains instrument B only.

Chart 3.Attainable After-Tax Rate of Return

This setup permits a breakdown of the tax reduction (i*τ) into two components: the effective increase in the rate of return on savings (i*τ-ϕ), and transactions costs (ϕ). The maximum transactions costs incurred will be σa - σb. This also allows one to assess the cost of the distortion to preferences caused by the tax incentive, if these are defined as the maximum transaction costs that an investor will be willing to pay to reassert his preferences, after taking advantage of the tax break.

The effects of the introduction of the tax incentive are summarized in Table 5 in terms of (1) the increase in the rate of return on savings (which will increase the overall level of saving); (2) the influence on a final reallocation toward instrument B (which is only achieved if the original set of preferences is not re-established); and (3) the distortion caused (as here defined). The attempt to increase savings is unambiguously successful and a final reallocation toward instrument B will succeed, in spite of causing a distortion, if intermediation costs are sufficiently high relative to the distortion. In general, the smaller the distortion, the more successful is the policy in terms of increasing the savings rate and reallocating toward instrument B. However, there are costs to the policy, in addition to the foregone taxes, which are the smaller of the allocational distortion and transactions costs.

Table 5.Effects of Tax Exemption with Regulated Market
(i)Return on savings:i*τ - (σa - σb),
where ϕ > σa - σb.
i*τ - ϕ,
where ϕ < σa - σb.
(ii)Final reallocation toward instrument B:iff ϕ > σa - σb.
(iii)Distortion:σa - σb.

The Problem with Thresholds on Exemptions

In a setting similar to that above, but with a general tax exemption and without thresholds assigned to specific instruments, the investor can choose the instrument upon which to obtain the tax advantage. In these circumstances there will be no distortion caused by the tax incentive, no transactions costs will be incurred, and the full extent of the tax incentive will accrue to the rate of return on savings. This is outlined in Table 6.

Table 6.Effects of General Tax Exemption with Regulated Market and No Threshold to Exemptions
(i)Return on savings:i*τ
(ii)Final reallocation toward instrument B:no switching.

Distortions with Financial Liberalization

Financial liberalization will increase the range of instruments available to savers and these will be at a flexible and higher rate of interest. An analysis similar to that conducted above can be used to analyze the potential success and costs of an attempt to reallocate from a liberalized instrument bearing the higher rate of interest (instrument C) to the regulated instrument (instrument B). The own and relative rates of return on these instruments, with and without tax exemptions, are presented in Table 7(a) and (b) respectively. The tables present the own rate of return on the relevant instrument in the boxes on the downward-sloping diagonal and the relative rate of return from the instrument on the row, over that in the column, in the other boxes. This, by following the argument above, allows one to draw the following summary (Table 8):

Table 7.Relative Returns on Financial Instruments(Relative return on instrument in row over column)
(a). With Tax Exemptions for B and D.
i*(1-τ) + σaa - σb) - i-(i-i*)(1-τ) + σa - σcσa - σd + i*τ - (i - i*)
i*τ ∠ (σa - σb)i* + σbσa - σc + iτ - (i - i*)σa - σd - (i - i*)
(i - i*)(1 - τ) - (σa - σc)-iτ + i - i* - (σb - σc)i(1-τ) + σcσc - σd - iτ
iτ + i - i* - (σa - σd)i - i* - (σb - σd)iτ - (σc - σd)i + σd
(b). With No Tax Exemptions.
i*(1-τ) + σaσa - σbσa - σc - (i-i*)(1-τ)σa - σd - (i - i*) (1 - τ)
-(σa - σb)i* (1 - τ) + σbσa - σc - (i - i*) (1 - τ)σa - σd - (i - i*) (1 - τ)
(i - i*)(1 - τ) - (σa - σc)(i - i*)(1 - τ) - (σb - σc)i(1-τ) + σcσc - σd
(i - i*)(1 - τ) - (σa - σd)(i - i*)(1 - τ) - (σb - σd)- (σc - σd)i(1 - τ) + σd
Table 8.Effects of Tax Exemption with Financial Liberalization
(i)Return on savings:i*τ - (i - i*) + (σb - σc),
where ϕ > (i - i*) - (σb - σc).
i*τ - ϕ,
where ϕ < (i - i*) - (σb - σc).
(ii)Final reallocation toward instrument B:iff ϕ > (i - i*) - (σb - σc).
(iii)Distortion:(i - i*) - (σb - σc).

Note that the policy can again succeed in reallocating savings toward instrument B in spite of causing a distortion. However, with a larger increase in interest rates following financial liberalization (i - i*), the policy is less likely to work and the distortion is greater. Again, a final reallocation will only occur to the extent that transactions costs exceed the distortion. On the other hand, if the incentive to reallocate toward instrument C is sufficiently large relative to ϕ, substantial transactions costs will be incurred.

Distortions with Financial and Capital Market Liberalization

The effect of capital liberalization and market integration has been to severely limit the tax rate that can be imposed on the liberalized financial instrument. This is because it now competes with similar instruments abroad that may effectively bear only a limited withholding tax. As a result, a third goal of policy is now introduced, which is to limit the extent of financial intermediation of domestic financial assets occurring abroad. This was the case in France where exemptions were granted to SICAVs so that they could compete with similar foreign instruments. However, this has meant that the pre-existing system of granting tax exemptions has perverse results in terms of reallocating resources toward regulated domestic instruments.

If a financial instrument with a free interest rate and a tax exemption (instrument D) is introduced, for the purpose of limiting the flow of funds abroad, the effect of the tax exemptions on instruments D and B is to increase the relative preference for instrument D over the regulated domestic instrument B. As is evident from Table 7, the relative preference for instrument D with tax exemptions [(i - i*) - (σb - σd), from Table 7(a), box 14] is greater than without tax exemptions [(i - i*)(1-τ) - (σb - σd), from Table 7(b), box 14]. Thus, the system of tax exemptions has become counterproductive as regards allocating toward regulated domestic instruments, and the exemptions reduce the likelihood of a reallocation toward instrument B. These results are summarized in Table 9. The distortion introduced between these instruments amounts to τ(i - i*) - (σb - σd) and increases with the tax rate. However, it is lower than for a liberalized financial instrument with no tax exemption (instrument C) because both instruments (B and D) are given the same tax treatment.

Table 9.Effects of Tax Exemption with Capital Liberalization
Return on savings:τi* - (i - i*) + (σb - σc),
where ϕ > τ(i - i*) - (σb - σc).
where ϕ < τ(i - i*) - (σb - σc).
Reallocation toward instrument B:iff ϕ > τ(i - i*) - (σb - σc).
Distortion:τ(i - i*) - (σb - σc).

More specifically, the types of returns from saving principally considered are deposit interest, interest on bonds, dividend income, and capital gains on share investments.

The system of taxation of savings in France was recently described by a prominent French banker as “defying logic” (Le Monde (1993)).

While such observations are instructive, the value of cross-country comparisons of systems of taxation of returns from savings can be limited, and can lead to assessments that are less than conclusive. This is because financial and capital taxation systems across countries differ with the perceived goals of domestic policies (King and Fullerton (1983), pp. 307 and the following).

A general assessment of the optimal level of taxation of returns from saving, at an aggregate level, would be of limited value in the context of this review. Studies such as Razin and Sadka ((1989) and (1990)) have analyzed the optimal response to financial and capital market liberalization in terms of adjusting the aggregate level of taxation. However, taxation systems (and especially the French system) are seldom defined by uniform aggregate taxation rates and provisions. Also, domestic financial assets will typically differ in range and attributes and are not perfect substitutes for one another or for foreign instruments (Mintz (1990), pp. 355 and the following). For these reasons, it is deemed more appropriate to examine whether the composition of the system has adapted in a manner that is consistent with its goals, and the question of whether aggregate rates should be higher or lower is not addressed.

Capital liberalization in Europe raised the prospect of competitive tax reductions in different countries, as they tried to attract foreign funds, but attempts at establishing a uniform withholding tax rate have not been successful to date. See Gardner (1992).

In addition to M3-M1 (as previously defined), P2-M1 comprises Treasury bills and certificates, commercial paper issued by nonfinancial companies, long-term contractual savings plans, bonds, most (i.e., nonequity) mutual funds, and the compulsory reserves of insurance companies. Many of the tax exemptions on financial instruments are limited to specified amounts (as will be described below) and Chart 2 is constructed on the basis that these limits are not exceeded.

While these data are an indication of the extent of the reliance on labor income, to the exclusion of income from savings and other sources, a complete review of the taxation systems involved will be required fully to establish this point. This is because the OECD data cited involve broad categories that will, for some countries, include some taxation on labor with “other personal income.”

In addition to the instruments listed in Table 4, a number of government bond issues have, from time to time, been exempted from taxation.

This limit will not apply, however, to funds transferred to specified long-term mutual funds or invested in bonds to finance privatization (bons Balladur) before the end of 1993.

The Belgian Prime Minister has indicated his intention to try to establish a universal withholding tax of 15 percent in Europe (L’Echo (1993)).

    Other Resources Citing This Publication