Journal Issue


International Monetary Fund. European Dept.
Published Date:
February 2019
  • ShareShare
Show Summary Details

Reforming Italy’s Social Welfare System and Lowering Taxes on Labor: some Considerations2

The Italian authorities are keen to improve Italy’s social welfare system. They plan to introduce a new citizenship income program, aimed at assisting the poor and facilitating their reintegration into the labor market. They also plan to gradually lower taxes on labor income. With a social safety net centered on pensions and the elderly and an array of income support schemes as well as a heavy tax burden falling on labor income, there is significant scope to improve Italy’s system of taxes and transfers to promote growth and inclusion. This note uses the tax-benefit microsimulation model for the European Union (EUROMOD) to evaluate possible reforms to modernize Italy’s social safety net as well as ways to lower the tax wedge on labor.

A. Introduction

1. Italy has been struggling with relatively weak economic and social outcomes that in part reflect gaps in the social welfare system and a high tax burden on workers. Unemployment is high at about 10 percent but, in the South, is twice the national average and, among youth, is three times the national average and among the highest in Europe. The share of the population at risk of poverty is not only higher than the European average but is especially so among younger households.3 Real incomes of employees and other workers remain below pre-euro accession levels, as are the real incomes of younger generations and middle-aged households. On the other hand, the real wealth and incomes of older households and pensioners remain above the levels of two decades ago. While these outcomes largely reflect the relatively weak growth performance of Italy over the past two decades, the relatively favorable treatment of the elderly at the expense of the working age and younger populations reflect a fragmented social safety net that is centered on generous pensions, a heavy tax burden falling on labor income, and high structural unemployment rates.

At Risk of Poverty

(Percent of population)

Source: Eurostat.

2. This note evaluates possible reforms to modernize Italy’s social welfare system and lower the tax wedge on labor. Potential spending increases (from modernizing the safety net for the poor) and revenue reductions (from lowering taxes on labor) relative to the current baseline would need to be considered, alongside a reform of design features. However, it is important to note at the outset that these measures would need to be fully consistent with a fiscal consolidation path that ensures Italy’s public debt declines firmly. In other words, complementary measures are critically needed to lower current spending and broaden the tax base. Tackling these challenges is best achieved as part of a comprehensive reform package to boost productivity, investment, and job creation; see Andrle and others (2018a) for details. The rest of the paper is organized as follows. Section B provides an overview of Italy’s welfare system, considers principles for reform based on cross-country good practice, and simulates the cost of a modern safety net using EUROMOD. Section C similarly provides an overview of Italy’s personal income tax regime, simulates the cost of moving toward a flatter regime that has previously been proposed by the government coalition, and discusses some key reform principles. Section D concludes.

B. Reforming Italy’s Welfare System

B1. Overview of the Welfare System

3. The Italian social welfare system is concentrated on pensions, inadequately supports children and people in working age, and is complex and fragmented. Social protection spending is the fifth largest in Europe, accounting for approximately 30 percent of GDP. The bulk of it is in pensions, reflecting a high share of elderly population and generous pension benefits— see Andrle et al. (2018b). High and rising pension spending has, over time,4 crowded out resources for other priorities, such as public investment, and redistributed resources away from children and persons in working age.5,6 Correspondingly, non-pension social benefit spending is lower than the euro area average, such as for social inclusion, family/child benefits, and housing.7 Numerous programs aim to provide social insurance and assistance, but point to complexity, fragmentation, and potential overlaps (e.g., in the various income support schemes at different levels of government)—illustrated briefly in the table below.

Composition of Social Protection Spending

(2014; Percent of GDP)

Source: Eurostat.

A Complex and Fragmented Social Welfare System

I. Social Insurance

  • A. Benefits related to the reduction in working ability or end of the working activity

    • 1. Old-Age, Invalidity, Disability, and Survivors’ Pension

    • 2. Mobility, collective dismissals, and unemployment benefit (NASPI and DIS-COLL)

  • B. Benefits related to the temporary suspension of the working activity

    • 1. Redundancy payment—Cassa Integrazione Guadagni (CIG)

    • 2. Salary supplement for agriculture workers

    • 3. Sickness leave

  • C. Family and social allowances

    • 1. Marriage leave

    • 2. Family Allowance

    • 3. Maternity Allowances

    • 4. Family leave for severe disability

    • 5. Inability Allowances

  • D. Other benefits

    • 1. Compensation benefit

    • 2. War Pension

II. Social Assistance

  • E. Family Support

    • 1. Family Allowance for families with at least three children

    • 2. Bonus for Babies (to be replaced by REI in 2020)

    • 3. Allowance for the birth or adoption of minors

    • 4. Day-care Bonus

    • 5. Scholarships and Free Textbooks Supplies

  • F. Anti-poverty allowances

    • 1. Social Pensions and Social Allowances to individuals older than 65

    • 2. Assegno di Disoccupazione, or ASDI (to be replaced by REI in 2018)

    • 3. Minimum Insertion Income (discontinued)

    • 4. Social card (partially to be replaced by REI in 2018)

    • 5. Support for Active Inclusion (to be replaced by REI in 2018)

    • 6. Inclusion Income (REI)

    • 7. Housing Benefit

    • 8. Young Culture Card

  • G. Benefits related to the reduction in working ability

    • 1. Civil Infirmity Allowance

    • 2. Monthly Assistance Allowance

    • 3. Accompany Benefit

    • 4. Frequency Benefit

    • 5. Sightless Pension

    • 6. Special Benefit

    • 7. Deaf-Dumb Pension

    • 8. Communication Benefit

    • 9. Personal, Long-term Assistance Allowance

  • H. Other benefits

4. Although Italy is among the largest spenders on social benefits, the coverage of the poor is among the lowest. In Italy, 27 percent of the poorest 20 percent of the population do not receive any benefits while a further 34 percent receive inadequate benefits. Consequently, Italy excludes about 60 percent of its poorest population from adequate social protection, while the euro zone on average excludes about 40 percent. The most vulnerable categories are working-age families with children. By contrast, retirees are covered by the pension system and have seen their poverty rates declining over time. Reform priorities, therefore, need to focus on poverty alleviation.

Coverage of Social Protection Benefits in the Poorest Quintile

Sources: World Bank (2016a), “Coverage, Adequacy, Gaps and Duplications of Social Welfare Benefits in Greece. Interim paper prepared for the Greece Social Welfare Review.”

* Includes simulated family benefits othewise not fully covered in the HBS (2014).

5. An important gap in the social welfare system is the lack of an extensive two-pillar unemployment benefits system or a last-resort but adequate income support scheme. This leaves unprotected a large fraction of the non-working poor. Individuals of young age are among the most vulnerable. In this sense, the social welfare system has weak anti-poverty properties.

  • In many countries, a two-pillar unemployment benefits system is implemented to protect the unemployed and non-working poor. A first pillar is for those who have worked and contributed to the unemployment insurance fund. A second pillar is for those who have either not contributed or exhausted their unemployment insurance benefits (see Ravagli 2015 for details). Examples of countries with an extensive two-pillar system include Germany, Finland, Ireland, and the United Kingdom.

  • In Italy, the second pillar exists only for those who have exhausted their unemployment insurance benefits (other examples are Austria, France, Portugal and Spain). It is not available for those who have not contributed or for the non-working poor (Box 1).

  • Italy introduced in 2017 a centralized last-resort income support scheme for the non-working poor, called the Inclusion Income (REI). But at a budgetary cost of €2 billion per year (0.1 percent of GDP), it is inadequately funded and only partially supports the population in poverty. Under full take-up, it would reach 45.8 percent of households in absolute poverty and 22.5 percent of those in relative poverty (Baldini and others, 2018). It was aimed at gradually replacing various income support schemes that were mostly designed and implemented at the local level (Box 2). It provides a transfer to households experiencing harsh economic conditions, i.e., based on a double means test of household income and wealth.8 The benefits are distributed via an electronic card and the amounts vary between a minimum of €187.5 for a one-person household and a maximum of €485.41 for households with 5 or more members.9 To obtain the benefit, the applicant must participate in a personalized work/social program that involves all civic services (e.g., municipalities, job centers, schools, and healthcare).

  • Against this backdrop, for the non-working poor, Italy needs either an extensive second pillar unemployment benefits scheme or an adequately funded and well-designed last-resort income support scheme.

Box 1.Italy: Recent Developments in Improving Unemployment Benefits

The 2015 Jobs Act introduced a more comprehensive and harmonized unemployment benefit, NASPI – Nuova Assicurazione Sociale per l’Impiego. Those who are entitled to this claim are workers who have lost their jobs and who have paid their national insurance for at least 13 weeks in the previous four years or 18 days in the last 12 months.1 Alongside the introduction of NASPI, the Jobs Act introduced two additional schemes: (i) an unemployment insurance for dependent self-employed workers (DIS-COLL); and (ii) an experimental assistance scheme (i.e. Assegno di Disoccupazione or ASDI) for workers who are no longer entitled to NASPI, have children, or are close to retirement age.2 ASDI was replaced by the Inclusion Income (Reddito di Inclusione—REI) in 2018.

In Italy, except for ASDI, all other types of unemployment benefits are attributed on a contributory basis; therefore, large groups of non-working poor remain unprotected (if unemployment assistance is picked as the instrument of choice to alleviate poverty).

1 Receivers of NASPI are entitled to 75 percent of their average salary of the previous 4 years, up to a maximum of €1,195 per month. As an incentive for people to seek employment, the NASPI payment will decrease by 3 percent every month after the fourth month. The NASPI can be claimed for half the number of months the claimant has paid their national insurance for, up to a maximum of 2 years. Entitlement to NASPI is conditional on claimants’ active attempts to seek employment and to their participation in programs of professional training.2 This benefit will be equal to 75 percent of the last NASPI payment and available for a maximum of six months.

Box 2.History of Poverty Reduction Measures in Italy

Designing and implementing poverty-reduction policies has largely been delegated to local governments, with nationwide programs tailored mostly toward the elderly (e.g., social pensions) and those with disabilities. This has left a large share of the population, especially the young and children, weakly protected. One of the early efforts to tackle poverty in Italy was the Minimum Insertion Income (RMI) that was introduced in 1999 on an experimental basis. RMI was supposed to replace the plethora of existing social assistance programs and be conditional on activation programs. Boeri and Perotti (2002) argue that while the intention of RMI was good, the implementation was dismal and little was learnt from its experimental nature. As an emergency measure to provide limited support to low-income families affected by the global financial crisis, the government introduced a social card in 2008, which was subsequently re-designed and broadened in scope in 2011 based on a mix of cash transfers and social services.

The government launched the SIA (Support for Active Inclusion) program in 2013, targeting low-income families with children/disabilities and in a limited number of municipalities. With the 2016 Stability Law, Italy set up, for the first time, a Fund for Combating Poverty and Social Exclusion with the aim of introducing a minimum-income program at the national level by 2018. In the meantime, the SIA program was extended nationwide in 2016 and its eligibility requirements were relaxed in 2017. SIA and other pre-existing income support measures will gradually be replaced by the Inclusion Income program (REI). The program aims to tackle poverty, re-organize welfare services, and improve coordination among social services.

6. In the 2019 Draft Budget Plan, the government announced the introduction of a Reddito di Cittadinanza (RC or Citizen’s Income Program) that is expected to replace the REI. The design elements remain under discussion and are to be specified in a forthcoming legislative intervention. Several variants for the RC scheme have been put forward. Among the issues are:

  • target population (e.g., full coverage of poor Italian citizens), means testing (e.g., there could be a test for income and wealth), other constraints on eligibility (e.g., beneficiaries cannot refuse more than two consecutive job offers), and the period of benefits (e.g., for up to three years);

  • size of the benefits (e.g., a monthly payment up to the relative poverty line of €780 per eligible citizen), possibly based on a top-up formula (i.e., an individual who pays rent10 and has zero income is entitled to the full amount, while incomes of others will only be topped up to the relative poverty line);

  • incentives related to participation in formal labor markets; and

  • conditional on a service pact for reintegration into the labor market, active and documented work research, attendance at training courses, and availability of useful work for the community.

7. The authorities will also need to consider the transition from REI to RC, taking into account the costs of RC (the amount budgeted in the November 2018 Draft Budget Plan was ½ percent of GDP per year), implementation capacity, and risks of poverty traps and those related to beneficiaries’ incentives to look for formal work and pay taxes. Given the prevalence of tax evasion and the size of the shadow economy, great care will be needed to ensure that incentives are not distorted. Distortions could arise from a high participation tax (i.e., if benefit withdrawal rates are steep, once a beneficiary becomes employed) and because some beneficiaries could try to combine the receipt of the benefits with informal work. A strong income effect (from generous benefits) might also encourage low-income workers to stop looking for jobs and fall into poverty traps.

B2. Reform Principles for a Guaranteed Minimum Income (GMI) Program

8. A well-designed and adequately-funded guaranteed minimum income program, or GMI, provides the most transparent instrument to support the non-working poor. GMI is, typically, a selective, conditional, and means-tested scheme. It is selective because it targets the poor. It is conditional because recipients must prove their commitment to finding a job or participating in activation programs (e.g., employment and training). It is means-tested because the entitlement depends upon the amount of income and wealth owned by the household. By comparison, a basic (citizenship) income scheme tends to be universal as it applies to all citizens regardless of their socio-economic status or their need, unconditional as it is granted to individuals without the need to meet any requirement, and not means-tested.

9. While GMI programs differ across countries, they are often aimed at poverty reduction and share a following general features:

  • Coverage and duration: working-age individuals and their households who have fallen into poverty by guaranteeing a minimum income. Benefit duration is unlimited. Most countries provide additional housing allowances, health care etc. Credible means-testing is essential.

  • Benefit level: the “guaranteed” household income limit is usually in the range of 40–70 percent of the “at-risk-of-poverty” 11 level to avoid poverty traps. The state tops up the income of the beneficiary up to that limit.

  • Incentives: well-designed schemes incorporate features to incentivize work, such as income disregard12 or gradual benefit phase-outs.

  • Conditionality: minimum income provision is generally conditional on participation in active labor market policies (ALMPs); eligibility is also linked to asset criteria.

The next four paragraphs elaborate on these features in turn.

10. GMI programs should aim for broad (or universal) coverage with credible means testing and not be time-bound.

  • Modern welfare states support individuals/households in poverty. Below the relative poverty line, the benefits should be universal in coverage (i.e., not restricted by geographical or demographic criteria) and not bound by a specific time limit. In this regard, note that both REI and RC appear to be time bound. REI has a duration of 18 months (extendable for another 12 months with a 6-month gap), while RC could have a duration of up to 3 years.

  • The ability to verify households’ income and assets is crucial to determine eligibility and benefit levels. This is best achieved by a credible means test that, although administratively burdensome, can substantially reduce errors of inclusion or exclusion. A credible means test allows GMI programs to rely on a top-up formula whereby the state tops up beneficiaries’ incomes up to a minimum guaranteed income level. REI demonstrates that Italy has sufficient capacity to implement accurate means tests. Nevertheless, given the high level of informality, appropriate verification of financial assets and income declaration of the self-employed should be considered in the design of the RC to avoid errors of inclusion.

11. The benefit level or the guaranteed income should reflect minimum needs without causing welfare dependence. To achieve this balance, the “guaranteed” household income limit is usually set in the range of 40–70 percent of the “at-risk-of-poverty” level to avoid poverty traps. In 2016, the at-risk-of-poverty threshold for Italy was €9,748 per year or €812 per month. Simple benchmarking indicates that the monthly benefit would be in the order of €325–€568. By this metric, if the RC benefit level is set at €780 per month, it would stand out as generous whereas the REI with €3,100 annual or €260 monthly income limit would not cover the basic needs.

GMI Benefit as a Share of the Poverty Threshold in the EA, 2015–16

(family with two adults and two children, in percent)

Sources: Eurostat, MISSOC Comparative Tables, and IMF staff calculations. Data on poverty thresholds refers to 2015 (2014 for Ireland); GMI levels refer to MISSOC January 1, 2016 update. GMI excludes other benefits (e.g. family and housing allowances).

12. Program design should include features that incentivize work. Generous benefit levels (the income effect) and high withdrawal rates (i.e., the reduction in benefits once beneficiaries find jobs, which is called the participation tax or the price effect) could strongly disincentivize work and discourage labor force participation. To strengthen incentives, successful GMI schemes in practice incorporate conditional in-work tax credits (including for secondary earners) as well as a variety of “out of work” benefits, such as (marginal) income disregard for part-time and casual work, gradual benefit phase-outs, and back-to-work bonuses. REI, however, has a high withdrawal rate (though below 100 percent). If RC has a top-up benefit formula that, absent other tools, also creates a high withdrawal rate, it risks falling short of its objectives of facilitating work and exits from poverty traps.

13. Participation in ALMPs is essential. To receive GMI benefits, all able-bodied adults should be required to register with the public employment services and be available to participate in activation services. This would further reduce disincentives to work and control the costs. Traditionally, conditioning benefits on participation in public works has the desirable feature of self-selecting out those with informal employment. Conditioning on job training or placement, education etc. would help the households to move to employment. But if there is a high degree of welfare dependence (such as in a poorly-designed scheme), ALMPs are likely to be less effective.

14. A pragmatic way forward would be to gradually expand REI as the basis of a modern GMI scheme in the context of a comprehensive review of Italy’s social welfare system. REI has desirable features such as the use of credible double means test (based on the ISEE indicator, see footnote 8), a functioning delivery infrastructure, and centralized administration. However, its coverage is limited as it is unable to reach all households in relative poverty, the benefit levels are below international benchmarks, it is time bound, withdrawal rates are relatively high, while the local administration of ALMPs is under developed.13 Building on REI’s desirable features, a pragmatic option would be to gradually scale up REI into a modern GMI in the context of a comprehensive review that rationalizes the existing array of income support schemes. Additional resources could be directed to REI to ease the eligibility criteria and scale up benefit levels to international benchmarks, and to reinforce the mechanisms in REI that attenuate disincentives to work. The modern GMI scheme should include in-work tax credits for low-wage secondary earners as well as gradual benefit phase-outs and income disregards, which would reduce scope for abuse. Further work on upgrading the effectiveness of ALMPs would also be needed.

B3. EUROMOD Simulations of the Cost of Reform

15. This subsection simulates the fiscal cost of a hypothetical GMI program in Italy using EUROMOD.14 The poverty threshold adopted, the “at-risk-of-poverty”, is 40 percent of median equivalized household disposable income. Transfers would top up households’ disposable income to this poverty threshold based on double means tests of household income and wealth (ISEE below €6,000 and the income component less than €4,000). Disability benefits are fully disregarded. All other social benefits, including family allowances, are treated as income. Furthermore, the household must not have real assets (excluding the main dwelling) worth more than €20,000 and financial assets worth more than €10,000. Perfect targeting is assumed with the benefit awarded to all eligible households (i.e., full take up). It is assumed that there are no increases in taxes and social contributions and/or reductions in other social benefits.

16. Covering about 1.8 million households, such a GMI program would cost about €9 billion (½ percent of GDP). The number of beneficiaries is estimated at 7.1 percent of Italian households, corresponding to 4.1 million individuals, which is 6.9 percent of the total population. The average benefits to households would be about €4,800 per year or €400 per month. Benefits would be strongly concentrated in the South where poverty levels are higher.

17. By comparison, simulations undertaken by the Italian Social Security Institute (INPS) estimate the cost of a potential RC program at €12 billion per year. The simulations are illustrative of costs related to different design considerations. INPS conducted a technical analysis using a tax-benefit microsimulation model based on stricter eligibility conditions but higher benefit levels: (i) household income and wealth (ISEE) below €9,360; (ii) household must not have real assets (including the main dwelling) worth more than €30,000 or financial assets worth more than €6,000 for single-person households, €8,000 for couples, and €10,000 for three-person households;15 (iii) benefits are calculated based on a top-up formula; (iv) welfare transfers are fully subtracted; and (v) assuming that there is a 50 percent under-declaration of financial assets.16 Such a hypothetical RC program would cover more than 2 million households, or 4.3 million individuals, with average benefits to households of €6,000 per year. Close to 70 percent of the resources would be transferred to households in the poorest two deciles of the income distribution, and the benefits would be homogeneously distributed geographically. The withdrawal rate (participation tax or the price effect) would, however, range between 81–92 percent for different groups given its gradual phase-out design, adversely affecting incentives. This underlies the importance of ensuring adequately strong design to include features that incentivize work and avoid poverty traps.

C. Personal Income Taxation in Italy

C1. Overview of Personal Income Tax System

18. While the fiscal burden on labor income is high in Italy, tax rates are applied to a relatively narrow base. Labor income is subject to a progressive tax scale with a starting rate on the first earned euro of 23 percent17 and a top tax rate of 43 percent for income exceeding €75,000 (the personal income tax is known as “IRPEF”). Additional taxes are also due at the regional (1.23 percent to 3.33 percent) and local (0 to 0.9 percent) levels (addizionali IRPEF).

IRPEF for 2016
Taxable income bracket (€ per year)Tax Rate
≤ 1500023%
≥ 7500043%

High Tax Rates:

  • The implicit tax rate on labor18 in Italy (at 43.2 percent) is the second highest in the EU.

  • The share of personal income tax in total taxes is among the highest in the EU at 41 percent.

  • The progressivity of the effective marginal tax rate is steep between the second and third taxable income brackets owing to the increase of the statutory tax rate from 27 to 38 per cent and to the phasing out of the €80 in-work tax credit (see below).

  • The ratio of the social security contributions to GDP is 13.4 percent, which is 2 percentage points higher than the EU average.

  • The average tax wedge in Italy for a single person earning an average income is 47.9 percent, well above the OECD average of 35.9 percent.19 This pattern is observed across levels of income and types of households.20

Income Tax, Employee and Employer SSC (percent of labor costs)

Sources: OECD taxing wages database.

Relatively Narrow Base:

  • Italy offers a large set of tax deductions and tax credits on labor income (6 percent of GDP), see the table below.

  • The largest item is the ‘tax credits for income source’ (i.e., depending on different types of income, such as employment, self-employment, or pension income). This item reflects the fact that the first bracket of income (from zero to €15,000) is subject to a tax rate of 23 percent (i.e., there is no zero-tax bracket for low income). These tax credits are reduced progressively until, at income levels above €55,000, they no longer apply.

  • A National Income Tax Bonus (or an in-work tax credit) proportional to the annual income was introduced in May 2014 and modified by the 2018 Budget Law. The bonus is equal to €80 per month, which starts with a minimum annual gross salary of €8,000 and decreases proportionally for annual gross salaries between €24,600 and €26,600. The cost to the budget is about €9 billion per year (about ½ percent of GDP).

  • There are also family tax credits (including for dependent spouses and children) as well as tax credits for retirees, voluntary life insurance, education, mortgage interest payments, medical expenses, and refurbishment, among others.

  • The compliance gap on personal income taxes (i.e., the measure of tax evasion, or the difference between theoretical and actual tax revenues) is also high at about 2 percent of GDP (see left-hand side chart below).

  • Average taxes paid by gross income brackets and geographical areas are generally lower in the South and Islands (see IT-SILC, 2016), see right-hand side chart below.

Italy: Personal Income Taxes—Deductions and Tax Credits
Total taxable income817.3833.0843.0
Total deductions39.842.845.5
For main residence12.30.08.8
SSCs and other charges24.625.526.1
Net taxable income777.5790.1797.5
Gross tax (IRPEF)210.1214.2216.3
Total tax credit66.266.167.6
Income sources42.041.642.1
Refurbishment, energy savings, and rental fees5.96.37.2
Deductions at 19% and 26% or other tax credits5.25.45.6
Regional additional income tax (IRPEF)11.411.811.9
Net tax (IRPEF)151.2155.2156.0
Source: MEF and EUROMOD Country Report
Source: MEF and EUROMOD Country Report

Tax Gap

(Percent of GDP, average 2012–14)

Sources: MEF and IMF staff estimates.

Average Effective Tax Rate By Geographical Area and Gross Income

(2015, percent, income in euros, one earner)

Source: ISTAT.

C2. EUROMOD Simulations of the Cost of a Flatter Income Tax Regime

19. The authorities plan to gradually reduce the tax burden on both labor and business income. The 2019 Draft Budget Plan starts the process by extending the ‘regime forfettario’ for self-employed: setting the earnings threshold at €65,000 (previously €25,000–€50,000 depending on the activity), applying a 15 percent tax rate, and eliminating some requirements related to assets and personnel cost. As of 2020, a new regime will be introduced for self-employed with earnings between €65,000 and €100,000 taxed at 20 percent. The number of beneficiaries should increase from around 950 thousand to over one and a half million. The budgetary cost is about €1.8 billion (0.1 percent of GDP) at peak. The draft budget also introduces a permanent tax break on reinvested profits and changes several tax incentives (such as abolishing the allowance for corporate equity, introducing more stringent requirements for R&D tax credit, discarding the introduction of a dual tax regime for the self-employed and unincorporated businesses (IRI), and extending the hyper-amortization scheme).

20. Although not in the authorities’ draft budget plan, one possible tax reform is simulated, namely, a flatter tax regime on personal income, and found to be very costly and highly regressive. This possible reform was part of the coalition government’s program of policies. It is assumed that personal income tax rates are simplified to only two rates of 15 and 20 percent. Different income thresholds are used for when the rates apply (ranging from €30,000 to €80,000), with a tax allowance of €3,000 for low-income households. Unless accompanied by significant base broadening, i.e., maintaining the existing set of tax deductions and credits, such a reform would be very costly. The cost is estimated using EUROMOD at €80–90 billion (about 5 percent of GDP) depending on the income threshold. If the tax base were to be broadened somewhat, i.e., tax deductions and credits partly eliminated, the net revenue loss would be smaller. For instance, Baldini and Rizzo (2018) estimate that partial base broadening for a threshold of €80,000 on family income would result in a net cost of €50 billion (about 3 percent of GDP). The reform would be highly regressive, benefiting mostly the well-off in the top decile of the income distribution and resulting in a higher Gini coefficient (charts below).

Revenue Loss From Lower PIT

(Billion of euros)

Sources: Staffs estimates using EUROMOD.

Note:T=30,… T=80 refers to different income thresholds in thousand of euros.

Ineqality Implications of Lower PIT

(Change in Gini coefficient)

Sources: Staffs estimates using EUROMOD.

Note: T=30,… T=80 refers to different income thresholds in thousand of euros.

Disposable Income Under Different Taxation


Sources: Staffs estimates using EUROMOD.

C3. Reform Principles for A Comprehensive PIT Reform

21. Any consideration of lowering personal income tax rates should be in the context of a comprehensive review of the overall tax-benefits structure. It needs to be consistent with a fiscal consolidation path that ensures Italy’s public debt declines firmly (¶2). That is, before personal income tax rates are lowered, the tax base should be broadened with high-quality measures to ensure revenues in line with the recommended fiscal path. A comprehensive reform would: (i) maintain progressivity and support labor supply; (ii) ensure tax neutrality; (iii) enhance tax certainty; and (iv) broaden the tax base. The following four paragraphs elaborate.

22. Income redistribution considerations should be balanced with adequate labor-supply incentives. Statutory tax rate reductions could encourage labor supply but, as noted above, would be regressive, under the current distribution of marginal effective tax rates (across income percentiles, see text chart). Additional policy choices need to be considered:21

  • The effective marginal tax rate schedule should be progressive.

  • The large variability of effective marginal income tax rates that derive from regulatory stratifications (e.g., from the design of the €80 measure) distorts labor supply decisions. Statutory tax rates and income support programs should therefore be reshaped such that effective marginal tax rate hikes are smoothed. This would encourage labor supply.

  • Italy has the lowest labor supply of married women among EU countries driven in part by a tax credit for non-working spouses. Consideration should be given to better targeted measures for secondary earners, especially low-income ones. For instance, the family (“dependent spouse”) tax credit could be replaced with a tax credit for households if both spouses are employed. This so-called working family tax credit or in-work tax credit for secondary earners could be increasing with the number of children. The available evidence suggests that adopting in-work tax credits for low income earners has sizable positive impacts on female labor-force participation and aggregate employment (Saez, 2002; De Mooij, 2008). For Italy, Colonna and Marcassa (2015) find that replacing the dependent-spouse tax credit with an in-work tax credit would increase the married-women participation rate by 3 percentage points.

Italy: EMTR Across Income Distribution

Sources: Staffs estimates using EUROMOD.

Note: Income percentiles are on the horizontal axis.

23. Reforms should ensure tax neutrality, so that decisions are based on economic merits and not tax reasons. The authorities should be mindful that the new fiscal plans could depart from neutrality across legal forms and business income. Currently, dividend income, which is subject to corporate income taxation (at 24 percent), is charged a final withholding tax of 26 percent, yielding a combined tax rate of 43–43.8 percent. This is close to the top marginal personal income tax rate of 43 percent that applies to profit withdrawals by sole proprietorships under the dual tax regime for unincorporated businesses, or IRI. It, thus, has maintained neutrality across legal forms and business income. The 2019 Draft Budget introduces lower effective tax rates for artisans, small businesses and professionals, and plans to cancel the IRI regime. All else equal, reducing the corporate income tax rate to 15 percent and the top personal income tax rate to 20 percent (applied also to sole proprietors) in the case of full implementation of the flatter tax regime would result in a combined tax rate of about 37 percent for dividend income, compared to a top marginal personal income tax rate of 20 percent. Harmonizing tax rates across different types of investment income is advisable. This example further illustrates the need for a systemic approach to income tax reform.

24. The frequency of tax changes is relatively high in Italy compared to other G20 countries and contributes to tax uncertainty. Renewing temporary measures with varying conditions is prevalent, which can generate uncertainty when their expiry date is either unclear or not credible. Such uncertainty risks creating a “hold-up problem”, where firms under-invest until the uncertainty is resolved.22

Frequency of CIT Changes (1983–2014)

Source: IMF staff estimates.

25. High quality measures are needed to broaden the tax base. Addressing large compliance and policy gaps, rationalizing tax expenditures, stricter enforcement, and introducing a modern tax on primary residences based on updated cadastral values would help broaden the tax base. On average and over the period 2013–15, the tax gap–which measures the extent of tax evasion, or the difference between theoretical and actual tax revenue—was estimated by the Italian Ministry of Economy and Finance at around €109 billion per year, over 6 percent of GDP. The stock of unpaid tax and social security contribution debt in 2016 was €614 billion (about 35 percent of GDP), out of which only €31 billion (less than 2 percent of GDP) is deemed collectible. Permanent reductions in arrears require deep institutional (enforcement) and behavioral (compliance) changes for tax administration and tax payers, respectively. Recurrent tax amnesties could generate some quick returns but undermine the compliance culture and effectiveness of tax administration.

D. Conclusion

26. Italy needs a modern social safety net and a lower tax wedge on labor as part of a growth-friendly and inclusive fiscal consolidation package. This note used the tax-benefit microsimulation model for the European Union (EUROMOD) to evaluate possible reforms to argue:

  • Italy needs a well-designed and adequately-funded guaranteed minimum income program for the poor—one that is not time bound; caps benefits at levels that avoid welfare dependence; includes gradual benefit phase-outs, income disregards, or conditional in-work benefits to incentivize regular work; and has adequate controls to prevent abuse with effective local administrative capacity. It is recommended to introduce this program in the context of a comprehensive review of the social welfare system, by gradually scaling up the inclusion income program, at a cost of about ½ percent of GDP, and rationalizing other anti-poverty schemes.

  • Any consideration to flatten personal income tax rates should be undertaken in the context of a comprehensive review of the income taxes and the structure of benefits. It needs to be consistent with a fiscal consolidation path that ensures Italy’s public debt declines firmly, maintain progressivity and support labor supply, safeguard tax neutrality, enhance tax certainty, and broaden the tax base.


The author would like to thank Nazim Belhocine, Ernesto Crivelli, Csaba Feher, Daniel Garcia-Macia, Rishi Goyal, Shafik Hebous, Alvar Kangur, Zsoka Koczan, Davide Malacrino, Roberto Piazza, and seminar participants at the Bank of Italy, as well as the Italian authorities for helpful inputs, comments and suggestions.

The results presented here are based on EUROMOD version H1.0+. EUROMOD is maintained, developed and managed by the Institute for Social and Economic Research (ISER) at the University of Essex, in collaboration with national teams from the EU member states. We are indebted to the many people who have contributed to the development of EUROMOD. The process of extending and updating EUROMOD is financially supported by the European Union Program for Employment and Social Innovation ‘EaSI’ (2014–2020). We make use of microdata from the EU Statistics on Incomes and Living Conditions, Italy 2015 (_a2), made available by Eurostat/Istat. The results and their interpretation are the author’s responsibility.

The population at risk of poverty rate increased from 18.9 percent in 2008 to 20.3 percent in 2017 but is about 33.1 percent in the South and Islands.

While non-pension social spending is discretionary, pension expenditures are based on earned rights, which could take a long time to decrease owing to slow attrition of the pension system.

The Italian welfare state favors pensioners and male individuals in large firms and unionized sectors. The incidence of relative poverty and the in-work poverty rates are higher than the European average and grew during the crisis.

Ravagli (2015) shows that recipients of social pensions and supplementary pensions are frequently not poor and, in some cases, even belong to rich households.

Families have responded by relying more extensively on intra-family transfers for social assistance, although its scope is being reduced by declining family sizes; see Boeri and Perotti (2002).

The indicator of the equivalent economic situation, ISEE, must be below €6,000 and the income component must be less than €3,000 for eligibility. ISEE is an indicator based on four components: family income; real-estate assets; financial assets; and characteristics of the family unit. Furthermore, the household must not have real-estate assets (excluding the main dwelling) worth more than €20,000 nor financial assets worth more than €10,000.

The transfer is not a fixed amount and is computed as the difference between the ISEE and disposable income. From the latter, it is possible to deduce the amount of the rent paid, one of the most important expenditures for households in poverty. Since rents are generally higher in the North, this deduction enables part of the different cost of living between areas to be considered.

It is estimated that about 20 percent of the 5 million potential beneficiaries have their own homes, and the largest share is concentrated in the South. 46.5 percent of the possible beneficiaries are in the Center-North.

The “at-risk-of-poverty” rate measures the percentage of people out of total population with an equivalized disposable income below the poverty threshold, which is conventionally set at 60 percent of the national median equivalized disposable income.

For example, disability pensions are excluded or a part of income (e.g., from part-time and casual work, or dependent employment) is not considered in tests to avoid negative effects on labor supply. Part-time and casual work are the main sources of occupation for many GMI recipients. It is important to design the GMI in such a way as to encourage these types of occupation. A sizable income disregard would be an option.

Employment centers operate under the competences of different levels of government in Italy and the modalities of their cooperation with the private sector are not clearly defined, notwithstanding the creation in 2015 of the National Agency for Active Labor Market Policies (ANPAL). According to the Bank of Italy, in 2017, over 25 percent of job seekers had contacts with an employment center while the share of the unemployed who found a job in the private sector through employment centers was just 2 percent. Even in countries with a longer tradition of ALMPs, the probability that an unemployed person would find a job through an employment center is not high: it was 7 percent in France and Germany in 2016.

The model is static and does not account for behavioral responses to policy changes.

Increasing by €1,000 per additional family member.

If INPS were able to verify the financial assets declarations with the “national register of financial accounts”, the cost to public finances would be substantially lower.

Owing to the ‘tax credits for income source’, de-facto, employed labor income and pensions below €8,150 as well as self-employed incomes below €4,100 are not taxed.

The implicit tax rate on employed labor is defined as the sum of all direct and indirect taxes and employees’ and employers’ social contributions levied on employed labor income divided by the total compensation of employees.

The tax wedge is sum of taxes and social security contributions paid by employees and social security contributions paid by employers, minus family received benefits. The average tax wedge is the tax wedge divided by the total cost of labor for the employer. This measure can be computed at various levels of income and types of households (singles, couples, with or without children).

To reduce the labor tax wedge, Italy has adopted several measures, including social security contribution exemptions, the €80 bonus on personal incomes taxes, and deduction from the regional production tax (the so called IRAP) base of the labor cost of hires with permanent contracts.

A complementary measure to reduce the labor tax wedge would be to consider reducing the employers’ social security contribution rate closer to the EU average (see Andrle et al, 2018b). Given wage rigidities and being in a monetary union with major trading partners, it could reduce labor costs (and producer prices, including those of exports) and increase labor demand in the short term. However, the impact on employment and output may be longer lasting if it is accompanied by a shift in the tax burden toward non-labor income (VAT and property taxes) that is less distortionary. Reductions in employers social security contributions should further be coordinated with pension policies since, over time, these would lead to a decline in pension benefit outlays through the Notional Defined Contributions scheme.

Gulen and Ion (2016) present evidence that policy uncertainty is persistently and negatively correlated with corporate investment, with an important part of the negative effect of tax-related uncertainty measured as the presence of temporary measures where the expiration date or the possibility for renewal is unclear.

Other Resources Citing This Publication