The Distributional Impact of the Fiscal System in Albania

In a context of fiscal consolidation and the need to deliver on a structural reform agenda, policy makers in Albania must not lose sight of the critical redistributive role of the fiscal system, particularly its impact on poverty and inequality. Using household survey data, this paper estimates the redistributive effect of fiscal policy on income distribution and poverty in Albania, assessing the individual and combined effects of taxes and public social spending. The findings show that the fiscal system in Albania plays a positive role in reducing inequality. Yet, it has a moderate poverty-increasing effect. Specifically, taxes and social protection contributions have a poverty-increasing effect; indirect taxes, particularly the value-added tax, account for the largest increases in poverty. This effect is somewhat compensated by direct government transfers, which are pro-poor and equalizing, but are not large enough to offset fully the negative impact on the taxation side. Ongoing reforms aimed at improving the efficiency and targeting of social assistance can contribute to enhancing the pro-poor impact of the fiscal system.

In a context of fiscal consolidation and the need to deliver on a structural reform agenda, policy makers in Albania must not lose sight of the critical redistributive role of the fiscal system, particularly its impact on poverty and inequality. Using household survey data, this paper estimates the redistributive effect of fiscal policy on income distribution and poverty in Albania, assessing the individual and combined effects of taxes and public social spending. The findings show that the fiscal system in Albania plays a positive role in reducing inequality. Yet, it has a moderate poverty-increasing effect. Specifically, taxes and social protection contributions have a poverty-increasing effect; indirect taxes, particularly the value-added tax, account for the largest increases in poverty. This effect is somewhat compensated by direct government transfers, which are pro-poor and equalizing, but are not large enough to offset fully the negative impact on the taxation side. Ongoing reforms aimed at improving the efficiency and targeting of social assistance can contribute to enhancing the pro-poor impact of the fiscal system.  The CEQ analysis usually excludes some categories of taxes and spending such as corporate income taxes, defense, spending on infrastructure, and other public goods due to the difficulty of assigning the benefits or burdens to individual households. This is more likely to be true for the poverty analysis, which depends on absolute incomes, than inequality, which depends on relative incomes.
Our findings show that the fiscal system in Albania plays a positive role in reducing inequality, while it has a poverty-increasing effect. The fiscal system contributes to reducing inequality from a Gini coefficient of 0.35 at market income (pre-fiscal income) to 0.32 after all taxes and transfers are considered. However, in the baseline scenario of pensions treated as pure deferred income, it also contributes to increasing poverty, going from 7.5 percent to 8.1 percent ($2.5-day poverty line). Specifically, while direct and indirect taxes and social contributions have a role in reducing inequality, they also have a povertyincreasing effect; indirect taxes, particularly the VAT, account for the largest increases in poverty. This effect is somewhat compensated by direct transfers (with programs such as the Ndihma Ekonomike) which are pro-poor and equalizing, but not large enough to fully offset the burden placed on the less well-off.
The rest of the paper is organized as follows. Section II describes the fiscal system in Albania, providing the context for this paper. Sections III and IV describe the methodology, as well as the application to Albania, respectively. Results are presented in sections V and VI, including the impact of the fiscal system on poverty and inequality, with a subsequent zooming into the role of taxes and transfers. Section VII showcases some additional uses of the tool for policy making, by simulating changes in the personal income tax, and Section VIII concludes.

II. Overview of the Fiscal System in Albania
The structure of the fiscal revenue in Albania relies heavily on indirect taxes, as well as direct taxes and social contributions. Table 1 shows that the Value-Added Tax (VAT) represents nearly 33 percent of the government total revenue and 9.1 percent of GDP. When compared to the other taxes collected from tax offices and customs, VAT accounts for about half of the revenue. In addition, the excise tax revenues represent 11 of total revenue and about 3.1 percent of GDP. Direct taxes and contributions, namely the Personal Income Tax (2.2 percent of GDP), and Social Insurance and Health Insurance (4.8 percent of GDP), also provide an important source of revenue to the government, representing 25 percent of overall government revenue. if they have an excess credit (on the VAT paid for the inputs). Also, there are various products that are exempted from this tax and for which firms cannot reclaim the VAT paid for the inputs.
 Excise: Excise duties is an indirect tax per unit (per liter, kilogram) levied on manufactured and specific imported goods that are considered to cause "harm" or are luxury goods. The main categories of goods taxed are fuel, alcoholic drink and tobacco and cigarettes.
 Social and Health Insurance Contributions: Employed persons and their employers shall be liable to pay social and health care insurance contribution. In calculating social insurance, Albania relies on a system of minimum and maximum gross salary boundaries (ALL 22,000 and ALL 97,030). The employee's contribution rate is 9.5 percent while the employer pays 15 percent. The contribution rate for health care insurance is 3.4 percent on the monthly gross salary, with the employer and the employee paying equal shares, respectively 1.7 percent and 1.7 percent. On the other hand, the self-employed pay 23 percent for social security and 3.4 percent for health insurance, but for the latter the contributions are calculated on the double of the minimum wage. While these contributions are compulsory for all economically active persons, resident in Albania, other non-active persons can voluntarily pay the social and health contribution.
On the spending side, social protection, education and health are among the largest categories. Overall expenditures in Albania amounted to 32.8 percent of GDP in 2015, with a large share of spending under the social protection category amount 9.12 percent of GDP (Table 2).  Energy Subsidies: The beneficiaries entitled to the energy compensation are families that receive social assistance (Ndihma Ekonomike), disability pension, old-age pensioners, paraplegics and tetraplegics, and low-income families with a monthly salary below 35,000 ALL/month. The compensation received by these families in 2015 is 1,288 ALL/month (approximately 0.1 percent of GDP), from which 648 ALL/month is benefited as compensation for the removal of the protective threshold category of the energy consumption of 300kW/month and 640 ALL/month is received as compensation for the increase in the electricity price (up to 200 kWh per month).
 Education (in-kind transfer): The majority of pupils and students attending school are enrolled in the public system. In the academic year 2014-2015, the public education system accounted for: 90 percent of enrollment in primary schools, 89 percent of enrollment in secondary education, and 85 percent of enrollment in post-secondary education (Ministry of Education and Sports of Albania, 2016). The first two levels of education are compulsory and free. Both central and local governments are responsible 9 for these levels: serving pupils from preschool to the ninth grade (primary education) and serving students from tenth grade up to the twelfth grade (secondary education). As to the third level of education (bachelor, master and doctoral studies), spending is carried out only by the central government. The total spending on education in 2015 was 2.80 percent of GDP, a share with low variation during the last decade. This account for 8.5 percent of total government spending.
 Health (in-kind transfer): The health system in Albania is mainly public, both for primary and hospital services. In-kind health benefits are offered to: (i) the insured population that pays compulsory health contributions, (i) self-insured persons who either are formally self-employed or voluntary pay contributions, (iii) economically inactive persons, whose contributions are paid by the state budget or other categories defined by the law 9 (children below the age of 18, old-age pensioners, persons receiving social assistance etc.), and (iv) the uninsured population who may wish to access the public health system, by direct out-of-pocket payments from those individuals. The total amount spent on health in 2015 reached 2.84 percent of GDP, totalling 40,984 million ALL. This accounts for 8.7 percent of total government spending.
Further details of these subsidies and transfers are provided in Annex 4.

III. Methodology
To assess the distributional impact of the fiscal system in Albania, the analysis follows the Commitment to Equity Methodology (CEQ) developed by Lustig (2017). The methodology is centered around defining a set of income concepts that include or exclude specific fiscal interventions (Figure 2), to evaluate the impact of the fiscal system and specific fiscal interventions on poverty and inequality. How much does the fiscal system contribute to changing market income inequality? Does it contribute to reducing poverty? Which taxes and transfers are progressive and/or pro-poor? What would be the distributional impact of changing a fiscal instrument? These are some of the questions that are answered in the paper.

The analysis relies on and calculates the following key income concepts:
 Market income (or market income plus pensions) includes factor income such as wages and salaries, income from capital (rents, profits, dividends, interests), private transfers (remittances and other private transfers such as alimony, etc.) before taxes, and social security contributions.  Market income plus pensions, equals market income plus the subsidized portion of the income from contributory pensions.  Net market income, constructed by subtracting direct taxes and contributions (personal income taxes and employee contributions to social security) from market income plus pensions.  Disposable income, adds direct transfers to net market income, thus including a combination of added transfers and subtracted direct taxes and contributions.  Consumable income (or post-fiscal income), subtracts indirect taxes (VAT and excises) from disposable income.  Final income, adds in-kind transfers to consumable income, namely social spending on health and education.
The treatment of contributions to pensions and income from pensions in a country may differ depending on the design and functioning of the pension system. There is no consensus in the literature on how to treat pensions as it depends on the design and functioning of the pension system in each country; normally two scenarios are constructed: a benchmark case in which pension income is treated as pure deferred income, and thus included in market income; a second (sensitivity) scenario in which pension income is treated as a pure government transfers, and pension contributions as a direct tax.
The analysis relies on typical indicators of a standard incidence analysis to unpack the distributional impact of the fiscal system, 10 which include commonly-used measures of progressivity, poverty and inequality:  Concentration coefficient: The coefficient of concentration (or quasi-Gini) is an index summarizing the concentration curve of a tax or transfer, and it ranges between -1 and 1. The curves from which this measure is derived depict the cumulative percentage of households (from poor to rich ranked by market income) on the horizontal axis and the cumulative percentage of tax (transfer) paid (received) by each centile. This is used jointly with the pre-fiscal (before any taxes and transfers are accounted for) income Gini to construct the Kakwani index.  Kakwani index: A useful summary statistic to measure progressivity. The Kakwani index for taxes is defined as the difference between the concentration coefficient of the tax and the Gini for prefiscal income; for transfers, it is defined as the difference between the Gini for pre-fiscal income and the concentration coefficient of the transfer. A Kakwani index for taxes will be positive (negative) if a tax is globally progressive (regressive). A Kakwani index for transfers is positive if a transfer is progressive in relative terms. 11  Redistributive effect: it captures the marginal contribution of the net fiscal system element(s) to the Gini coefficient of inequality. The marginal contribution is understood as the difference between the Gini coefficient with and without the tax or transfer. If positive, it captures a redistributive effect, so a decline in the Gini.  Poverty reduction effect: it captures the marginal contribution of the net fiscal system element(s) to a poverty headcount defined at a certain poverty line. Again, the marginal contribution is understood as the difference between the poverty rate with and without the tax or transfer. If positive, it captures a poverty reduction effect, so a decline in poverty.
It is worth noticing that a progressive tax is not necessarily equalizing (positive redistribution effect) or poverty reducing (positive poverty reduction effect). 12 Also, the net fiscal system can be equalizing but impoverishing. 13 Therefore, measures of progressivity need to be combined with marginal contributions to assess the effect of specific instruments on poverty and inequality to get a full picture.
10 For additional information see Lustig and Higgins (2017). 11 A tax is globally progressive if the proportion paid in relation to pre-fiscal income increases as income rises. This is the case if the concentration curve lies everywhere below the pre-fiscal income Lorenz curve. A transfer is globally progressive in relative terms if the proportion received in relation to pre-fiscal income decreases as income rises. This is the case if the concentration curve lies between the pre-fiscal income Lorenz and the 45-degree line. 12 The first phenomenon is known in the literature as the "Lambert" conundrum (Lambert, 2001;Lustig and Higgins, 2017). Taxes, for instance, can be regressive according to the Kakwani index but when combined with transfers make the system more equalizing than without the regressive taxes. For a thorough discussion see, for example, Enami, Lustig, and Aranda, 2017. 13 See Higgins and Lustig (2016).

11
The methodology focuses on first-order effects, meaning it does not account for behavioral responses to changes in the net fiscal system. It is also assumed that the economic incidence of direct taxes and contributions are borne entirely by the income earner, and the burden of indirect taxes is borne by consumers.

Household Survey Data
The analysis uses data from the 2015 Household Budget Survey (HBS) for Albania, which provides information on the socio-economic situation of the Albanian households. The HBS is a continuous survey of private households covering the entire territory of Albania, representative at the urban and rural level, and at the prefecture level. It collects data both from a 14-day diary and a direct interview. In 2015, it interviewed 7,335 households. The 2015 HBS 2015 combined with macroeconomic and fiscal data from national income accounts for 2015. The section below on assumptions details the shortcomings of the HBS data for this analysis, and how these were addressed.
Although the HBS reports income data, its overall quality, particularly for labor income, is low. Disposable income seems to be largely underreported in the HBS, as shown by a consumption aggregate 14 that is 2.5 times higher than the reported disposable income. Labor income is collected at a household level and not at an individual level, and estimates of labor income per employed for households with employees in the formal-sector employees are much smaller than estimates from administrative data. 15 These estimates suggest a large underreporting of labor income, likely linked to the high informality rates and a large agricultural sector in Albania. Given this, the analysis follows recommendations (Lustig 2017) of using consumption instead of income as a basic measure for incidence analysis when income data are unreliable. In this case, the starting point is to set disposable "income" equal to the household consumption identified in the survey. From there, the method goes backwards to estimate market income (removing direct transfers, and adding direct taxes and contributions), and forward (removing indirect taxes and adding in-kind transfers) to estimate consumable and final income, respectively.
The analysis is complemented by data from national accounts, administrative information for certain programs and design features of each specific intervention.

Fiscal Interventions Included
On the revenue side, the analysis includes the personal income tax (PIT), social security contributions, VAT and excises. These account for 78 percent of the total tax revenue and 70 percent of the total fiscal revenue. Therefore, the analysis does not include other taxes that affect households, such as custom tax and property taxes (at 0.4 and 0.3 percent of GDP respectively). Although it does not directly include taxes paid by firms (profit tax or VAT by firms), it is assumed that, in equilibrium, firms pass the VAT tax burden to consumers (full pass-through).
On the spending side, the analysis includes social protection spending, as well as health and education spending. These account for 45 percent of the total spending and 62 percent of the total classified spending 13 in 2015. Other spending categories (defense, public order and safety) and are likely to be less relevant from a distributional impact perspective.

Empirical Strategy
The empirical strategy is described below, including the assumptions made for the allocation methods used to match household-level data with fiscal and program information, and to work around the challenges encountered given the microdata at hand.

Direct taxes
The application of the method to Albania includes personal income taxation and social security contributions. It assigns the statutory rates for wage and self-employed to those formally employed (informality assumption explained below) by grossing up labor income. It assumes that the burden of taxes paid by the employer falls fully on the employees. 16 The application of the methodology to Albania makes assumptions of labor market informality, which implies non-payment of personal income tax or social security contribution. Those assumed to be formal in the labor market and contributing to PIT and SSC are private sector employees with a permanent job or indefinite time contract, or temporary job with short term contract; or employees of the public sector or international organizations. The informal workers include the private sector employees in temporary jobs without a contract, occasional or seasonal job, and unpaid workers. The informality rate derived from this assumption (among the employed 15-65) is 33 percent in 2015, which is comparable to 31.4 percent in the 2015 Labor Force Survey. 17 Two important challenges were present in identifying individual labor income. The first challenge was the identification of the portion of household disposable income (in this case equivalent to total household consumption) that was attributable to labor market income. Given that labor income data in the survey was assessed to be largely underreported, some steps were taken and assumptions made to identify labor income: (i) direct transfers and market income from other sources, as reported in the survey, were subtracted from total consumption; (ii) the residual, labor income, was scaled-down at the household-level and for all households. The adjustments were made across the board, using as a scale down factor the matching of formal-sector wages in the survey with the corresponding wages in administrative data.
The second challenge relates to assigning the total labor income identified to employed members earning income within the household (thus excluding unpaid family workers), given the lack of individual-level wage or earnings data for each household. This is important in Albania given the progressivity in personal income taxes. Data showed that a large share of households with earners had only one employed household member (57 percent), in which case all labor income (from wage or self-employment) was assigned to that person. If more than two household members worked (35 percent) and were wage employed, labor income was assigned using ratios for main and second earner estimated from the Labor Force Survey 2013, conditioning on educational levels. 18 If two or more people in the household were self-employed (with only a few of these cases in formal employment, thus relevant to taxation), it was assumed that this was a family business so all but one member were assigned the minimum wage, and one person was assigned the remainder subject to a self-employment flat tax rate. 19 For households with one wage and one selfemployed, the model uses ratios from the income data reported in the survey between these two income sources to assign it to each of the two household members. For the cases in which more than two peopled worked (8 percent), labor income was equivalently assigned in per capita terms.

Pensions and social protection spending
The methodology for Albania uses the old age pension income received from pensions in Albania, as well as the income from several social protection programs -including the two main social assistance programs in Albania, as reported in the household survey.
Old age pensions are treated as savings/deferred income and thus part of market income, as in many other countries around the world that have conducted this type of analysis. Given the contributory nature of pensions and their size relative to market income in Albania but, most importantly, their unique lifecycle aspect -with people contributing throughout their working lives to receive income in old-age -recording them as pure Government transfers may not be accurate. Nevertheless, as explained below, results for both scenarios are presented.
The survey captures information on other sources of income treated as transfers. This includes disability pension, survivor pension, unemployment benefits, social/care services for the elderly and disabled, and Ndihma Ekonomike (NE, the poverty targeted social assistance program). Even though, by design, some of these are contributory benefits, their contributions and benefits are usually treated as taxes and transfers in these types of analyses given their shorter-term nature and people's perceptions of these as such (as opposed to savings). Nevertheless, given their small size with respect to households' total income, results are not expected to be influenced by this methodological choice. These categories do not match perfectly with the full spectrum of benefits available in Albania, some quite small in coverage and benefit, so households might be attributing the source of certain benefits to others. Nevertheless, the main benefits -disability pension and Ndihma Ekonomike -are presented separately in the survey and likely more accurately reported. Due to data limitations, the disability income cannot be disaggregated into disability assistance (non-contributory) and disability pension (contributory). Given that the number of beneficiaries of NE in the survey is significantly underestimated when compared to administrative data, a propensity score matching method was applied, following the methodology by Souza, Osorio, and Soares (2011), 20 resulting household types: A) Households in which both members have tertiary education; B) Households in which no member has tertiary education; C) Households in which one member has tertiary and the other does not. Then, the estimated coefficients are used to allocate labor income in the different household types. In A) The main earner received around 10 percent higher labor income than the second earner. In B) The premium is about 20 percent, while in C) this is 19 percent. We focused on tertiary since previous evidence for Albania showed that the wage premia is larger for tertiary. 19 This assumption implies that only the owner is subject to self-employment taxation rules. Small businesses in Albania are usually owned by one person and, although all family members work in this business, they are usually treated as wage employed and registered with the authorities as such. The latter rule is strengthened and better implemented from 2015 onwards given the increased efforts of the government against informality. In addition, there are incentives from the household side to register other household members as minimum-wage salary workers under the family business, as they would be 0% PIT rate instead of selfemployment taxes. 20 This method is commonly used to adjust for underestimation of beneficiaries of a particular program. The method consists in imputing beneficiaries that did not report to be recipients of the NE program in the survey, but are "likely beneficiaries". First, we run a Probit model of program participation against household per capita consumption, possession of various household assets and 15 in 87 percent of the beneficiary households recorded in administrative data to be identified in the survey, 21 and a total 9.3 percent of all households in Albania as beneficiaries of this social assistance program in the survey. For the disability pension benefit, 41 percent of beneficiary individuals recorded in administrative data are identified in the survey. 22 Targeted energy transfers are not captured in the HBS, so they are simulated into the data. We do not have information on beneficiaries and benefits received, so we estimate the latter based on the program rules. In particular, we use the eligibility criteria of the benefit to identify beneficiary households and assign the monthly benefit amount to the family (1,288 new leks per month). Households are eligible if they either (i) receive poverty targeted cash transfer (Ndihma Ekonomike); (ii) have household heads who receive a disability pension, but without other employed household members; (iii) have household heads who receive an old age pension, but without any income from employment, and/or working age members in the household; (iv) have government workers receiving a salary of less than 35,000 ALL, but without other employed household members.
We assumed that all of those who are eligible receive the subsidy. This assumes perfect targeting and takeup. Despite this assumption, the simulated number of beneficiaries in the survey is smaller than in administrative accounts (approximately 153,965 and 203,928 respectively, still capturing 90 percent).
Assuming program leakage and imperfect take-up will lead to an even higher underestimation of beneficiaries.

Indirect taxation
Indirect taxes are simulated using data on households' reported consumption of the corresponding items. The team applies the VAT rate to the consumption item of the households. However, given the high evasion in Albania on payment of VAT, many consumers may not directly pay indirect taxes. We can either model tax evasion by assuming those who purchase in certain areas are tax evaders, or by assuming effective tax rates that reflect the rates paid in reality. 23 Given lack of quality data on place of purchase, we assume an effective VAT tax rate of 14 percent (compared to the 20 percent statutory VAT rate). 24 This assumption was informed by existing analysis of VAT efficiency 25 and by data on VAT revenue collection. Total VAT collection captured in the household survey represents 54 percent of total VAT collection in national consumer durables, number of children and other sociodemographic variables. Then, we randomly sampled households out of the beneficiary households and match these beneficiary households to non-beneficiary households with the closest propensity scores. Program benefits of NE are then imputed to the matched households, where the amount of benefit imputed is equal to the amount received (reported in the survey) by the household's matched beneficiary household. The idea is so match the number of beneficiaries in the survey to the national accounts as closely as possible. However, in our case, we do not match exactly the number of beneficiaries in administrative data due to some restrictions with the method: namely, the number of beneficiaries who need to be randomly sampled and matched to non-reporters has to be lower than the number of beneficiary households. The original number of recipient households in the survey was 38,588. Applying the method, we simulate around 69,568 beneficiaries, still less than the 79,339 recipients in administrative data (which represent about 10.4% of the population). 21 79,530 households in administrative data vs. 69,569 in the survey (with the simulation).
accounts. However, given that reported consumption in the household survey is lower than consumption in national accounts, the most appropriate validation is that total VAT collected in the survey as a ratio of total household consumption in the survey, is in line with the same comparison at the level of national accounts (VAT revenue vs private consumption). The analysis does not include VAT paid by businesses.
For excises, the exercise captures the direct effect of excises (i.e. excises paid directly by households when purchasing manufactured and specific imported goods subject to this tax), as well as the indirect effects that fuel taxes may have on products' prices. The latter is done using an input-output matrix for Albania in 2013 with 35 categories, to then map household consumption expenditures from the HBS to the I/O production sectors. The excise is applied by measurement unit of the product, for which some assumptions were made in cases where the survey did not contain sufficient information to differentiate between products. This was the case, for example, for excises of beer which depend on the alcohol content, not observed in the survey. Based on local knowledge, it was assumed that beer contained on average 5 percent alcohol content. Similarly, on fuel spending, the survey does not contain information on quantity consumed but only on total expenditure. Quantities are estimated using average prices based on the monthly reports of the Ministry of Energy. In terms of evasion, the compliance gap for excises is estimated to also be large, as in VAT, and particularly for road fuels and cigarettes. 26 There usually is, however, also underreporting of this type of consumption items in the household survey (alcohol, cigarettes and others), pushing down the amount of VAT collected. As such, the exercise does not alter the tax rate to adjust for evasion as with VAT, and does not aim at identifying evasion at a broader level, given underreporting in the survey.

Social Spending: Education and Health
On education, the survey does not identify school enrollment, neither in public nor in private schools. However, alternate sources of data indicate that school enrollment is high 27 and that access to public education is widespread in Albania. For instance, 98 and 97 percent of students in basic education and secondary, respectively, attend public schools. Most tertiary school students also attend public universities. 28 To identify beneficiaries of public spending on education, the method identifies age ranges by education system levels. Across education levels up to secondary, the strategy assumes universal coverage except for households who report education expenses on private education. 29 For those in tertiary and also excluding those who report expenses on private education, tertiary enrollment is assigned to those who are identified as students in the labor market module of the household survey and older than eighteen. Robustness checks compare the number of students in each level of the education system with administrative data of number of students enrolled in the public education system; results show that survey information is relatively close to that in administrative records. 30 Once we identified potential beneficiaries, we impute education spending among those identified as beneficiaries, the method uses both central government per capita spending, as well as prefecture- capita spending on education, by level of education. To avoid overestimating the redistributive impact of education (and health, as per below) spending -given that the per capita spending assigned comes from national accounts data while this is not the case for other components of the fiscal system in the model-, the value of per capita spending is scaled-down. 31 On health, we allocate in-kind health benefits by identifying individuals who may be using the different types of health services and impute the (scaled down) value received from public health services. 32 The scaled-down value of prefecture-level per capita in-patient and out-patient health spending, obtained from fiscal data, plus a distributionally-neutral element (as the difference between the national account value and in/out-patient costs, representing the spending by beneficiary on administrative costs, recurrent spending, and health system investments) is assigned to eligible individuals in the household survey, using explicit beneficiary criteria. Beneficiaries include those contributing to health insurance, as well as categories identified as subsidized beneficiaries. 33 In total, 2.03 million persons in Albania are identified in the survey as beneficiaries of in-kind health benefits (out of 2.89 million in the population in the 2015 HBS). Public dental care and health care in schools are excluded in the analysis because they are seldom used by the public and not covered by the health insurance scheme. For the scaling down of public health benefits, we use the same criteria we use to scale down education benefits.
Notice that this approach to impute in-kind transfers ("production cost approach") does not consider variance in the quality of services provided nor reflect different valuation of these services across the welfare distribution.

Model performance
The strength of the model and its assumptions are assessed by comparing the ratio of each fiscal element's total value to total private consumption in the survey, with the same ratio from national accounts. In other words, it is important to assess whether the relative size of the fiscal element in the economy is equal or close to the relative value of that element represented by the HBS. These checks are presented in Table A1, alongside results comparing how much of each element is captured by the survey (the share of the total amount in the survey to the revenue/transfer in the national accounts).
These checks show that across elements, and particularly the ones with larger weight in the fiscal system, the model performs relatively well. Moreover, the results are balanced between taxes and transfers, with the model capturing 64.2 percent of the value of those taxes in national accounts, and 64.8 percent of the value of included transfers. 31 The spending is scaled down so that the ratio of total education (and health) spending to total consumption from national accounts equals the ratio of social spending to consumption in the survey (Chapter 6, Higgins and Lustig, 2017). 32 We have a variable in the HBS capturing household health expenditures in the last 3 months (by public/private), but we decided not to use it to identify individuals who are actually using health services due to the short reference period. 33 According to the law on Compulsory Health in the Republic of Albania", compulsory health insurance contributions and payments shall be obligatory to all economically active persons with permanent residence in Albania, including employees, selfemployees, unpaid family members, and other economically active persons. Compulsory health insurance covers, also, the following categories of economically inactive persons: persons benefiting from the Social Insurance Institute, persons receiving social assistance or disability payments, in accordance with relevant legislation, persons registered as unemployed jobseekers in the National Employment Service, foreign asylum seekers in the Republic of Albania, children below the age of 18, pupils and students below the age of 25 (provided they do not have income from economic activities), and other categories defined by specific laws.

V. Results: Distributional Impact of the Net Fiscal System in Albania
The net fiscal system contributes to reducing inequality in Albania. Before any fiscal intervention, the Gini coefficient stands at 0.37 and falls to 0.32 after the combined effect of taxes and contributions, and social spending. The largest reduction in inequality takes place from the effect of in-kind transfers (health and education). Direct taxes and transfers, and in-kind transfers (education and health) are equalizing. Indirect taxes (VAT and excises), in contrast, have an unequalizing effect but very small. Figure 3 shows the change in the Gini coefficient going from households' market income (which treats pensions as deferred income) to net market income (removing direct taxes and contributions), to disposable income (adding direct transfers to households), to consumable income (removing indirect taxes) and to final income (adding in-kind transfers). Box 1 considers an alternate scenario in which pensions are treated as government transfers instead of deferred income; nevertheless, given that pension system is currently primarily contributory, this scenario is not elaborated further. Poland (2014) Brazil (2008-09) United States (2011) Costa Rica (2010) Uruguay (2009) Croatia (2014) Iran (2011-12) Mexico (2010) Tunisia (2010) Chile (2013) Ecuador (2011-12) Colombia (2010) Russia (2010) Venezuela (2012) Bolivia (2009) Dominican Republic (2006-07) Tanzania (2011-12) Albania (2015) Paraguay (2014) Nicaragua (2009) El Salvador (2011) Peru (2009) Ghana (2012-13) Uganda (2012-13) Sri Lanka (2009-10) Honduras (2011) Indonesia (2012) Guatemala (2011) Jordan (2010-11) Ethiopia  The reduction in inequality is comparable to that of other countries. The 0.052 Gini points reduction is in line with results from other countries in the region such as the Russian Federation (0.063), but lower than countries like Croatia (0.092) and Poland (0.13) (Figure 4). 35 The impact is broadly in line with the size of the government in Albania from a cross-country perspective ( Figure A2.3), although with similarsized governments, countries like Chile, Uruguay and Mexico achieve a larger redistributive effect.
On the poverty side, however, the fiscal system was moderately poverty increasing in 2015. Focusing on households' absolute level of income, and using international poverty lines employed across countries in the CEQ sample of $2.5/day and $4/day (2005 PPP), our findings show a different picture from what is found on the inequality side. Starting from 7.5 percent of households in poverty at $2.5/day with pre-fiscal income, poverty increases slightly to 8.1 percent after the combined effects of direct and indirect taxes, contributions and direct transfers ( Figure 5 and 6). This increase is higher with the $4/day poverty line, going from 22.6 percent to 29.1 percent. Taxes and contributions, and particularly indirect taxes, play an important role in this poverty-increasing effect. In fact, the reduction achieved with direct transfers, which contribute to reducing poverty (as reflected by the change from net market income to disposable income) is more than offset by indirect taxes that households face. In a similar way, those already poor are made poorer ( 35 For Poland, see Goraus and Inchauste (2016). For Croatia, see Inchauste and Rubil (2017 The poverty effect shows a mixed picture across countries. In countries like Chile, Mexico, Russia and Colombia the fiscal system results in a reduction in income poverty. This does not mean that no poor person becomes poorer but that the number of people who move out of poverty is higher than those who become poor. In other countries, including Albania and Croatia, the number of people leaving poverty is smaller than those who become poor due to the burden of taxes. In other words, poverty increases from the combined effect of taxes, contributions and direct transfers. At the $4/day line, Albania stands out as one of the countries with the highest increase in poverty (Figures 7 and 8). Using different poverty lines, analysis from other countries in Europe, such as Croatia, Montenegro and Poland, also show sharp increases in poverty (at 70 percent, 62 percent and 51 percent, respectively, compared to Albania at 29 percent).
Our benchmark analysis focuses on the scenario of pensions as deferred income. Nevertheless, as robustness check, we also compare the difference in results between the different scenarios. Figure B1 and B2 show that excluding pensions from market income in the pure transfer scenario leads to a higher prefiscal inequality and poverty, highlighting the important role of pensions for households and their large size compared to other sources of market income. Not surprisingly, inequality and poverty drop sharply from net market income to disposable income, as pensions benefits (as direct transfers) are added. Interestingly, the poverty-increasing effect of the full fiscal system is not present in the scenario of pensions as transfers.
In this case, poverty declines from 23 to 8 percent ($2.5/day, 2005PPP), compared to 7.5 to 8.1 percent in the case of pensions as pure deferred income.

VI. Incidence, Progressivity and Marginal Contributions of Taxes and Social Spending
How each of the fiscal interventions contribute to the observed changes in poverty and inequality?
One way to answer this question is by examining the extent to which taxes and transfers affect household income across the welfare distribution. Another way is by looking at the marginal effect 36 of each fiscal component on the Gini coefficient and the poverty headcount. Intuitively, a tax or transfer has a sizable poverty and distributional impact if it is strongly targeted to those at the bottom of the distribution (as 36 The marginal effect is the change in market income plus pensions due to adding or subtracting only the given benefit or tax from market income plus pensions. captured by the concentration coefficients), and if it is large relative to their incomes (as captured by the size).
Zooming in to the elements of the fiscal system across the income distribution allows to better understand the distributional effects. Figure 9 shows the share of broad categories of taxes and transfers in pre-fiscal income (market income plus pensions), by decile. The shares for the bottom first decile are very large because the denominators (pre-fiscal incomes) in this decile are small. This decile of the distribution is a net receiver of social benefits, as shown by the positive net cash position. 37 Once in-kind health and education benefits are included, though, the net benefit for this decile is very large, increasing real incomes by about 40 percent. Contrarily, starting at the second decile, households in Albania are net payers into the fiscal system. Direct transfers represent a non-sizable share of household income outside the first two deciles, reflecting a weak targeting. In-kind health and education benefits are sizable for most deciles, declining gradually as we move along the income distribution. Direct taxes and contributions also represent a much smaller share of income in the bottom deciles, reaching 21 percent in the richest decile. The composition of taxes gradually changes from indirect to direct taxes as we move upward in the income distribution. In the next subsection, we take a closer look at the effect of specific fiscal elements summarized in Table  3. The analysis relies on the previously defined concepts of progressivity (Kakwani index) and marginal 37 The net cash position captures the difference between market income plus pensions and consumable income (equivalent to all payments of taxes and cash benefits), as a share of market income plus pensions. contributions to inequality (redistributive effect), and poverty (poverty reducing effect). The effect will also depend on the size of each fiscal intervention ( Figure 10) with respect to households' income.

Taxes and Contributions
Direct taxes and contributions in Albania are progressive with modest equalizing effects. The Kakwani index of progressivity is positive (indicating progressivity) and highest for the Personal Income Tax (PIT) (0.202), and positive but lower for health and social insurance 38 contributions ( Figure 11). With the exception of the PIT, the distributional impact is small, mostly due to the small size. Overall direct taxes and contributions result in a modest redistributive effect of 0.093 Gini points (Figure 12). Annex 3 provides a cross-country benchmarking of the progressivity of direct taxes in Albania; it shows that although smaller than a large share of countries with available data, it is in line with that of other countries in Europe.

Figure 12. Marginal Contributions of Direct Taxes and Contributions in Albania
Source: Own estimates using the Albania HBS 2015. Source: Own estimates using the Albania HBS 2015. Note: A positive value means that the contribution is in the desired direction (e.g., equalizing or poverty reducing) and a negative value is the contrary. The marginal contributions of the redistributive effect are estimated from market income plus pensions to final income; for poverty reduction effect, from market income plus pensions to consumable income (as per the literature).

The poverty increasing effect of direct taxes and contributions is small but not trivial (Table 3).
Poverty increases by 0.23 percentage points under the $2.5/day poverty line. The effect is larger for the $4/day line, with a marginal contribution to poverty from PIT of 0.87 percentage points and 0.91 for social and health insurance contributions (Figure 12), raising the poverty rate by 1.48 percentage points when combined. This is reflected in the overall picture of poverty changes across income concepts, in which poverty increased by this amount when moving from market income (including pensions) to net market income. The impact on poverty of direct taxes is linked to the relatively low thresholds for paying each of 38 Social insurance contributions include sickness, maternity, pension, accident and unemployment insurance. 24 the brackets of the PIT: if at least one member of the household earns more than $1.06/day (2005 PPP) in the formal sector (either wage or self-employed) that member pays a PIT of 13 percent. The threshold for paying a 23 percent PIT tax rate for the waged employed is $4.58/day (2005 PPP). Similarly, a person earning more than $0.80/day pays social and health insurance. As a result, around 9 percent of the population in Albania were impoverished (became poor or fell deeper into poverty at $4/day) after paying direct taxes and contributions, i.e. when moving from market income (plus pensions) to net market income (1.8 percent were impoverished at the $2.5/day poverty line).
Indirect taxes, on the other hand, are regressive and slightly unequalizing, and they result in sizable poverty increases. This effect is largely led by the VAT, given its larger size as a fiscal intervention. Specifically, the VAT has a negative Kakwani (Figure 13), and while its effects on increasing inequality are small (0.13 Gini points), it has a marginal contribution to poverty of 2.9 and 6.8 percent, with the $2.5/day and $4/day poverty lines, respectively ( Figure 14). In other words, the VAT pushed a share of households into poverty to an extent that direct transfers were not able to offset. Excises, although progressive, also have a negative impact on poverty. Around 29 percent of households were fiscally impoverished when moving from disposable to consumable income (i.e. after paying indirect taxes) at the $4/day poverty line (8 percent at the $2.5/day line). Compared to other countries in the region (see Annex 3) such as Armenia, Croatia, Georgia and Russia, indirect taxes were less regressive in Albania.

Direct and In-Kind Transfers
Direct transfers are all progressive in Albania and, to different extents, have positive redistributive and poverty-reducing effects (Figure 15 and 16). Looking at the set of transfers in Albania, with Ndihma Ekonomique (NE) and Disability Pensions as the relatively larger programs, 39 the positive Kakwani index shows progressivity across all, with NE as the most progressive program. They are also pro-poor, 40 as reflected in a negative concentration coefficient (Table 3), meaning that in absolute terms they are concentrated among the less well-off ( Figure 17). For example, around 53 percent of the NE benefits in 2015 go to the bottom quintile, and 75 percent go to the bottom 40 percent of the distribution. Leakage seems to be low, given that the households in the top quintile receive only 4 percent of the benefits. This reflects a pro-poor design, currently being improved through piloting of a proxy means testing in certain regions. Nevertheless, while most programs have positive but small impact given their size with respect to households' market income, NE and Disability contribute to reducing poverty the most by 0.8 and 1.2 percentage points ($2.5/day poverty line), respectively. Disability has larger poverty-impact despite not being a poverty-targeted benefit, which is driven by the larger amount spent: around three times the amount of NE.
Energy transfers in the other hand are not well targeted and represent a relatively small share of household budgets (low size), resulting in a marginal poverty and distributional impact. Unemployment benefits have a very low redistributive and poverty-reducing effect, mostly due to its small size (Figure 10), despite the fact that nearly 40 percent of the benefits go to the bottom quintile.

Figure 17. Concentration of social spending and transfers (by market income plus pensions, by quintiles)
39 Disability pension includes here both disability assistance and contributory disability pension. 40 A transfer is "pro-poor" if the proportion received in absolute terms decreases as income rises (globally progressive in absolute terms). Spending on education is progressive and equalizing, although quality and efficiency are not accounted for in the analysis. In particular, preschool and basic education are the most progressive education spending in Albania and also have the largest redistributive role, with a reduction in the Gini coefficient of 0.013 points (Figure 18 and 19). In fact, as found in other papers, the concentration coefficient of preschool/basic education is negative, suggesting that a larger spending share goes to the lower deciles, given that is disproportionately concentrated among the poor. Figure 17 above shows that overall education spending reaches people across the income distribution relatively equally, capturing the widespread use of public education services in Albania. As mentioned, these results do not capture heterogeneity in quality of services. The PISA results for 2015 show, for example, that a large share of 15-year-olds in Albania are considered functionally illiterate (50 percent with a score below 2 in reading), one of the highest among countries who take the test. 41 Moreover, within Albania there is a gap in PISA scores between urban and rural areas equivalent to, in some subjects, more than one year of schooling.
Similarly, health spending is progressive and equalizing. Health spending is more or less distributed across the population in absolute terms: About 20 percent of health spending goes to those at the bottom quintile of the distribution, while 15 percent is concentrated among the top (Figure 17). These differences may reflect high-income households opting out of the public system, Health spending contributes to reducing inequality by around one Gini point as it is slightly more concentrated among those at the bottom of the distribution, and its size is not insignificant.  Importantly, as with the majority of household surveys around the world, individuals with top incomes in Albania are not fully captured in the HBS. Therefore, the full extent of the policy impact cannot be evaluated, and results can be interpreted as a lower bound. 42 Analyses from other countries are exploring the use of alternative data sets, such as credit data in Indonesia, 43 to complement the household survey and capture the top of the income distribution.

The reduction in PIT, as measured in this paper, has a very small (simulated) impact on inequality.
The impact of the reduction in PIT on poverty and inequality across the different income concepts is shown in Table 4. As expected, the simulated PIT reduction is inequality-increasing, but the effect is very small, as shown by very small increases in the Gini index (measured using disposable, net, consumable and final). 42 Since the PIT is a progressive instrument, in the case of no-underreporting of labor incomes at the top, a rising PIT would have a disproportionately larger impact on richer households. 43 Ongoing work in Indonesia, as referenced in World Bank (2015) and World Bank (2018), is using the Central Bank's consumer credit database on mortgages, vehicle loans and others, to estimate top income households and combined, this with the household survey data to get a more complete distribution. The poverty headcount is reduced slightly, 44 but the simulated impact is very small. The results are robust to the choice of the poverty line. Simulation results show that such a PIT reform would reduce the redistributive effect of the tax, but the extent of the change cannot be observed using household survey data solely, as it does not include top incomes. Simulations show that the PIT becomes less progressive, captured in a slightly lower but still positive Kakwani index (Figure 20), and would have a smaller redistributive effect ( Figure 21). Not surprisingly, there is no difference in the poverty effect, as the change takes place in the top bracket of the direct tax. However, given lack of data on top incomes, the extent of the change cannot be fully explored. As previously mentioned, this observed effect is a lower bound and it would be expected that if top incomes were considered, the reduction in progressivity would be higher and the redistributive effect smaller.

VIII. Conclusions and Policy Insights
Distributional analysis provides an important tool for governments, particularly in reform processes.
It can contribute to understanding the household-level impact of fiscal policy, and how it can enhance the equity lens of its policy interventions or mitigate the effect when needed. This type of analysis provides one relevant perspective to fiscal system reforms, which should be accompanied by other considerations such as fiscal impact, and efficiency of revenue collection and spending.
This paper shows that the fiscal system in Albania plays a positive role in reducing inequality, while it has a slight poverty-increasing effect. This effect is observed in other countries in the region, although with different magnitudes. Direct and indirect taxes and social contributions have a role in reducing inequality, but they also have a poverty-increasing effect largely driven by the VAT. Direct transfers, with programs such as the Ndihma Ekonomike that are pro-poor and equalizing, partly offset this effect. Beyond providing a static picture of the fiscal system's impact, the tool allows for policy simulations. For instance, simulations around PIT reforms show that a reduction in the top bracket is expected to reduce the redistributive effect with no impact on poverty. Importantly, it must be noted that the poverty-increasing effect may be tempered by accounting for the positive welfare impact of public education and health services.
The analysis provides inputs to a policy discussion on several fronts. On the spending side, there is an important policy agenda to improve the coverage and targeting of direct transfers. Progress is being made in this area. For example, an improved design of the Ndihma Ekonomike, with a proxy means test formula to improve targeting, was piloted in three regions over the past years and is being implemented nationwide. With the poverty-targeting accuracy of Ndihma Ekonomike having improved since the implementation of the pilot and given that it is expected to continue improving as the new eligibility rule of the program reaches other areas of the country, the poverty-reducing effect of this program -already captured as the most progressive transfer in Albania -is likely to be higher, thus mitigating the current net effect of the fiscal system. Similarly, a pilot reform of the disability benefit is ongoing with the objective of revising the eligibility criteria to improve targeting towards the truly disabled and efficiency of the program. An important agenda also remains in terms of improving the efficiency of spending in health and education, for better and more consistent quality of service delivery across regions in Albania.
On the taxation side, there is an important and ongoing government agenda in Albania around improving tax administration and compliance. As weaknesses in these systems and processes are addressed to reduce informality in payment of taxes, it is important to understand the potential distributional impact of these measures to, if needed, mitigate it through complementary reforms. Considerations over adjusting the level or design of specific taxes should be accompanied by analysis of fiscal impact and efficiency.

This analysis has clear implications for a data and knowledge agenda in Albania.
On the data side, improvements can be made for a more systematic and disaggregated collection of quality administrative data at the local and national levels. On the household survey side, similarly, the design of data collection instruments and samples could be revisited to improve the alignment with the social protection system, for a better understanding of the role of the fiscal system for households' living standards. Finally, a difficult data agenda is pending to better capture top incomes in Albania. In other countries, this has been done using administrative tax records or banking data. It is important to explore which source of information could shed light on this issue for the case of Albania.
Systematically including this lens into policy making calls, therefore, for investments in building and strengthening the institutional setup to assess and monitor distributional impacts.

Personal Income Tax
A taxpayer of the income tax shall be a resident or non-resident person who generates income in the Republic of Albania from personal earnings, self-employment activity, property and property-based rights, capital, capital gains, banks or treasury bonds interests, rent, copyright and winnings from games of chance; Revenues exempt from the income: 1) Regulations on rights of disabled persons; 2) Child allowances and special assistance for supplies for newborn children; 3) Income received from insurance benefits from the social and health insurance scheme; 4) Economic benefits for individuals with no income or with low income; 5) Benefits received due to sickness, disasters; 6) Scholarships to pupils and students; 7) Income received, in cash or in kind, by the owners as a reward for expropriation made by the state for public interest; 8) Income exempted under international agreements ratified by the Parliament; 9) Income received as a result of financial compensation to former owners and former political prisoners; 10) Contributions of the employer for employee's life and health insurance; 11) Indemnity from final decisions of the court and specific compensation for court cost; 12) Income received from state institutions for achievements in science, sports, culture; 13) The transfer of the right of ownership over an agricultural land by a registered farmer to another farmer or a legal person who performs agricultural activity is exempt from personal income tax; 14) The contribution made by each member of a voluntary pension fund to the extent determined by the law on voluntary pension funds and contributions made by an employer or any other contributor on behalf of a member of the pension fund voluntary; 15) Return on investment, including investment income by the pension fund assets, during the administration of the management company.
The following persons are excluded from the income tax: 1) members of foreign diplomatic missions in Albania 2) members of consular representative offices 3) members of international organizations, who enjoy the status of a diplomat while performing their official functions in the Republic of Albania, in accordance with the international conventions or agreements signed or accepted by the Republic of Albania or the Albanian Government.

Tax base
(1) A tax base for the income tax of a resident shall represent all the taxable income generated during the tax period minus deductions.
(2) A tax base for the income tax of a non-resident shall represent the taxable income of the taxpayer generated during the tax period in the Republic of Albania.
Personal Income Tax for income from wages:

Deductions:
Taxpayers, who are not subject to the declaration of personal income, but want to fill the personal income declaration, with annual gross income from all sources up to 1,050,000 (one million and fifty thousand) ALL per year (indexed annually) but no higher than this amount, may be subject to declaration. For purposes of calculating the taxable income for these individuals, deductions are as the follows: 1) The amount of interest from the bank loan taken for education, for his/herself or for children and persons in custody; 2) Medical expenses for themselves or for children and persons in custody for the uncovered part of the compulsory health insurance (8,532 ALL, but not more than 32,000 ALL). Non-residents are not eligible for deductions of their taxable income.

Value Added Tax
The Value-Added Tax in Albania is levied at two different rates of 20% as standard rate, and 0% on special products 45 as displayed in the following Goods and services supply for commercial or industrial activities in the sea 0%

Excise Tax
The Excise tax is a tax applied on manufactured and specific imported goods, which are to be consumed in the Republic of Albania. This tax is calculated based on the measurement unit and the excise tax rates. This tax shall be levied on goods such as: a) Energy products; b) Alcohol and alcoholic beverages; c) Tobacco and tobacco products; d) Other products determined by the Law.

Product Excise
Coffee

Social Insurance Contribution
Employed persons and their employers shall be liable to pay social and health care insurance contribution. The contribution cannot be awarded under a minimum and over a maximum monthly wage. The employee's contribution is 9.5% while the employer pays 15.8%. Health insurance is levied at 1.7% for both employee and employer. The self-employed pay 23% for social security and 7% for health insurance. The types of insurance and the respective rates are presented in the following

Income basis used to calculate contributions:
1) The basis for calculating the contribution is the gross salary of the insured person. It is the salary appointed under the legislation in force, or in the employment contract, as appropriate, by time or volume of work and other supplements, of a permanent character, arising from labor relations, which cannot be below the minimum salary of 22 000 ALL. 2) The gross monthly salary, for the purpose of calculating social insurance contributions, from 01/01/2015 onwards, is: a) not less than the monthly minimum wage, equivalent to 22 000 (twenty-two thousand) ALL and up to 97 030 (ninety-seven thousand and thirty) ALL, for employed persons and self-employed persons; b) equal to the monthly minimum wage of 22 000 (twenty-two thousand) ALL for unpaid family workers. c) the monthly salary, for calculation purposes of the contributions to the state budget, for the period under transitional payment or early retirement pension of persons who perform constitutional functions or as senior state employees and for former soldiers and workers of State Police, Information Service, Republican Guard, Prisons Police, Firemen and Internal Control Service employees, is equivalent to the reference salary, according to their rank and function, at the time of termination of the employment contract. d) the monthly salary, for calculation purposes of the contributions of the state budget for persons who receive unemployment benefit, for former employees who have worked in mines, underground, and former employees of the military industry, benefiting special financial treatment, caregivers to paraplegic and tetraplegic and women who retire for the period of accomplishing the higher education, is equal to the minimum salary of 22 000 ALL.
Exclusions from the gross income for the purpose of calculating the social insurance and health care insurance are as following: 1) the benefits that the employee receives from social insurance funds; 2) the compensations from the price changes 3) in cash gifts, that do not have permanent and general characteristics; 4) the income from alimony (food allowance), determined by a court decision; 5) the value of antidotes, that the employer gives in kind to persons employed under the legislation in force; 6) the remunerations with non-permanent character, the rewards from the special fund for family disaster/distress, for persistent diseases, major natural disasters cases, and outstanding activities; 7) the amount of expenditures for employees who perform activities outside the residential center, which are given based on the legislation in force and that are not part of the salary; 8) the royalties or other similar compensation, which are not related to the current labor contract.

Category
Contribution Rates Contributions for pension insurance, labor contract> 87 hours per month (employee) not less than the minimum wage and no more than the maximum salary 9.5% Contributions for pension insurance, labor contract> 87 hours per month (employer) not less than the minimum wage and no more than the maximum salary 15% Contributions for pension insurance, labor contract< 87 hours per month (employee) not less than the minimum wage and no more than the maximum salary 9.4% Contributions for pension insurance, labor contract< 87 hours per month (employer) not less than the minimum wage and no more than the maximum salary 13.9% Contributions for pension insurance, labor contract< 1 week per month (employer) not less than the minimum wage and no more than the maximum salary 0.3% A self-employed individual, whose family members do not receive a salary, but work and live together with him/her legally, excluding the self-employed in agriculture, shall pay 23% of the minimum monthly wage of 22,000 ALL.

060 ALL
A Self-employed individual and the self-employed individual who employs other persons, excluding self-employed in agriculture, not less than minimum wage and no more than the maximum wage. For the unemployed persons, who receive unemployment benefits, the state budget pays contributions on minimum wage.

23%
For former employees who have worked in underground mines and former employees of the military industry, benefiting special financial treatment, caregivers to paraplegic and tetraplegic and women who retire for the period of accomplishing the higher education, contributions on minimum wage are paid by the state budget.

21.6%
For persons having constitutional functions and senior employees of the state, ex-servicemen, employees of the State Police Intelligence Service, the Republican Guard, Police, Prisons, Police, Fire Protection and Control Service Interior employees during transitional payment or early retirement pension, contribution are paid by the state budget on the respective average salary, according to their rank and function, which is then indexed annually based on the coefficients applied on the assessed basis.

Health Insurance Contribution
Liability to contributions: 1. Compulsory health insurance and the payment of respective contributions are mandatory for all economically active persons with permanent residence in Albania, as: for the persons employed with an employer, performing part-time and intermittent work assignments, self-employed persons with the status of natural person, owners of legal persons, who have senior functions in society, unpaid family workers, natural persons, self-employed in commercial or service activities, ambulant, who have reached the age of 16, self-employed in agriculture, capable of working, who have reached the age of 16 until retirement age, foreign nationals and stateless persons, employed, employers or self-employed in entities registered in the Republic of Albania.
2. The contributions payment for the persons benefiting from the Social Insurance Institute, persons receiving social assistance or disability benefits, in accordance with the relevant legislation, persons registered as unemployed jobseekers at the National Employment Service, asylum seekers in the Republic of Albania, children under the age of 18, pupils and students under the age of 25, provided they do not have income from economic activities financed from the state budget.

Income basis used to calculate contributions:
1. The basis for calculating the contribution is the gross salary of the insured person appointed under the legislation in force, or in the employment contract, as appropriate, by time or volume of work and other supplements, of permanent character, arising from labor relations, which cannot be below the minimum wage of ALL 22,000. 2. The state contribution for economically inactive persons is based on the healthcare consumption per capita, indexed by the rate of inflation. The consumption per capita of health care services is determined and approved by the Assembly, along with the annual budget approval. 3. The basis for calculating the voluntary health care contribution for self-employed workers, persons over the age of 18, with a regular income from movable property and real estate, self-employed in agriculture, shall be ALL 44,000. 4. The monthly minimum salary, for the purpose of calculating the health care contribution for selfemployed persons in agriculture, is ALL 4,620. 5. The monthly salary, for the purpose of calculating the contributions of the state budget for persons who have worked in underground mines and former employees of the military industry, benefiting special financial treatment, caregivers to paraplegic and tetraplegic and women who retire for the period of accomplishing the higher education, is equal to the minimum wage.

Contribution rates:
The contribution rate for health care insurance is 3.4% of the gross salary, according to the payroll, where the employer's part is 1.7% and the employee's part is 1.7%, but not less than the minimum wage. The self-employed and maintenance workers with the exception of selfemployed in agriculture, persons over the age of 18, with a regular income from movable property and real estate, self-employed in agriculture 3.4% 3.4%

Definition:
The Economic family allowance, known as economic assistance, is a (limited in time) support, in cash and in kind, to individuals with special status and to families in need.

Eligibility:
The beneficiaries of economic assistance shall be: a) Families without income or with insufficient income; b) Unemployed orphans over 25 years, who are not living in institutions or under foster care; c) Orphans aged 18-25 years, who are not settled to institutions of social services or under guardianship; d) Parents with more than three children born simultaneously, who belong to families in need; e) Victims of domestic violence and trafficking victims, for the period being under protection order or emergency protection order, who are not assisted by social care institutions.

Exceptions:
Families consisting of a single member, who meets the following conditions, are not beneficiaries of the economic assistance: a) He/she is the owner of any capital asset, with the exception of residential house and agricultural land; b) He/she is economically active (employer, employee or self-employed), except for occupationally disabled persons, blind persons, paraplegic and tetraplegic; c) It is abroad for any reason except for individuals who study or medical treatment as well as family members of persons appointed to work as diplomatic representatives of international organizations; d) It is not registered as an unemployed jobseeker, except for households living in the village, the invalids, the blind persons, paraplegic and tetraplegic; e) Rejects participation in work when it is provided by the employment office, participating in community service work which is organized by the city/municipality, and/or refuses to participate in vocational training courses being capable and of working age; f) He/she has refused to take land that has been given by Law No. 7501 dated 31.08.1991; Families living in villages and being in possession of agricultural land earn the right to partial economic assistance, except for families that: i. although they have been granted land under the law, they do not possess it because of land usurpation or ownership problem; ii.
They have usurped the land of someone else forcefully; iii.
They have sold and purchased or rented and leased the land The beneficiaries of the full economic assistance shall be those who do not receive income from: a) economic activity; b) assistance programs and social services or any other social insurance system; c) capital; d) family members who live abroad; e) being granted saline land; f) being granted land, which was later categorized as urban area for hydropower, water supply or any other activity.
The beneficiaries of the partial economic assistance shall be those who do not have receive sufficient income from: a) economic activity b) capital c) livestock, poultry, beehives, vineyards and gardening; d) pensions and other income, except for occupationally disabled persons, blind persons, paraplegic and tetraplegic, and parents of triplets or having more than three children born simultaneously; e) land according to the categories based on the production of field crop, vineyards, olive orchards, arboriculture (except lands being under desalination process), which are based on the following coefficients:

Benefit amount for areas except Tiranë, Elbasan and Durrës 47
Economic assistance shall be given in the form of monthly installments in lek, or in the form of monthly assistance in kind for the categories specified in Article 5 of this law. However, the letter is not used hitherto. a) 3200ALL per month for the spouses; b) 2 600 ALL per month for family members of working age and single spouses; c) 600 ALL per month for each additional household member of working age; d) 700 ALL per month for each additional household member below the working age (up to the age of 18). An additional amount of 300 ALL shall be given to those family members who attend the compulsory elementary school up. An additional amount of 100 ALL per vaccine shall be given to those family members who have been vaccinated according to the vaccination calendar.
The amount of economic assistance for parents with three, four and five (or more than five) children born simultaneously is 3 000 ALL, 4000 ALL and 5 000 ALL per month, accordingly. This economic assistance is received until the end of compulsory education but not more than 18 years. As to the orphans, the amount of benefit is the same (3000 ALL per month for each orphan). However, the amount of benefit, regardless of the family structure, cannot be more than 7 000 ALL per month.

Benefit amount for pilot areas only (Tiranë, Elbasan and Durrës):
a) 1 800 ALL for the first family member according to the family certificate; b) 1 260 ALL per month for the other family members being older than 18 years old; c) 900 ALL per month for the other family members up to the age of 18; d) 2 600 ALL per month in case of a family consisting of only one member.
The amount of economic assistance for parents with three, four and five (or more than five) children born simultaneously is 3000 ALL, 4000 ALL and 5 000 ALL per month, accordingly. The economic assistance for trafficking victims and domestic/family violence is 3000 ALL per month. However, the amount of benefit, regardless of the family structure, cannot be more than 8 000 ALL per month. 47 Tiranë, Elbasan and Durrës are considered as pilot areas.

Income Test:
Not applicable.

Benefit Amount:
Beneficiaries Benefit Amount: The widow/widower For the living spouse: 50% of the pension that a deceased had or should have had and 25% of the pension for any other member; Children/orphans If the living spouse does not benefit from the family pension, the children/orphans receive 50% of the pension that a deceased had or should have had; When the widow does not receive the family, but the beneficiaries are two or more persons, the amount of benefit shall be 25% for each of the family member eligible for the pension, but not more than 50% of the pension that the deceased had or should have had; If both parents die, the children/orphan receive a pension from each of the parents. The family pension shall not be higher than 100% of the pension of the deceased.

Early Retirement Pension
The early retirement pension for seniority (years of experience) shall be granted to military staff commissioned in the Armed Forces, State Police forces, Republican Guards, State Intelligence Service employees, Prison Police employees, Fire and Rescue Police employees, Internal Affairs Service employees. The amount of benefit shall be 12,264 ALL/monthly. These categories are eligible for other benefits and pensions such as compensation for electricity price change, bread and gas price change, income compensation for pensioners, provided that the sum of the early retirement pension and compensation(s) not be exceeded 14,414 ALL/monthly. Otherwise, the compensation(s) shall be partial until this amount is reached.