Tax Policy, Evasion, and Informality in Albania1
This note explores the factors underpinning Albania’s relatively low level of general government revenues. The analysis finds that while tax rates and tax expenditures are comparable to regional standards, tax efficiency is low and declining, perhaps due to relatively high levels of noncompliance and informality. Past revenue underperformance in Albania also reflects overly optimistic projections. To address the compliance and evasion issues, policies need to focus on strengthening the cooperation between the tax and customs departments, enhancing performance monitoring and governance of tax administration and improving policy design to improve tax compliance.
A. The Tax System in Albania: A Cross-Country Comparison
1. The size of the Albanian government is relatively small. General government revenues, both tax and non-tax, are lower in Albania than in its neighbors (Figure 1). Consequently, public spending is rather parsimonious.
Figure 1.General Government Revenue in Balkan Countries
Sources: OECD; IMF, WEO database; and IMF staff calculations.
* No breakdown between tax and nontax revenue available.
2. The tax burden is lower in Albania than in most other Balkan economies (Figure 2). Compared with most neighboring countries, the Albanian tax collections are modest for most of the main tax sources. In particular, income taxes, property taxes, and above all social security contributions are substantially lower in Albania. Only revenues raised by VAT and CIT approach levels that are in line with those of comparator countries. Therefore, Albania relies as much on indirect taxes as on direct taxes: the share of VAT and excise tax revenues in total tax revenues is about 49 percent.
Figure 2.Tax Revenue in Balkan Countries
Sources: IMF, WEO database; and IMF staff calculations.
3. The main headline tax rates in Albania are above the Balkan average, but the tax thresholds are also comparatively higher (Table 1). Relative to EU member states, Albania’s top rates for personal and corporate income taxes and social security contributions are lower than the EU average. On the other hand, relative to neighboring Balkan states, Albanian rates are rather high. However, zero-tax thresholds are higher in Albania than in its neighbors, on average. Also, because income tax efficiency measures are based on top rates rather than average, progressive tax regimes inherently measure as less efficient than flat rate regimes.
|Social Security Contributions|
|Bosnia & Herzegovina||25,524||17||10||10||10.5||31.0||41.5|
|Rep. of Kosovo||30,007||18||10||10||5.0||5.0||10.0|
|Rep. of Serbia||68,099||20||15||15||17.9||17.9||35.8|
|Unweighted Average Neighbors||29,015||18||12||16||13.0||17.7||30.8|
|Total Unweighted Average||22,098||19||17||31||16.6||16.6||33.2|
4. Albania’s lower tax revenues are partly a consequence of the lower efficiency of taxes on wages (Figure 3). Tax efficiency is the ratio between each tax revenues for each tax (in percent of GDP) and the top tax rate.2 Tax efficiency for social security contributions and personal income tax appears to be lower in Albania than in other Balkan countries. For corporate income and value added taxes, tax efficiency in Albania is roughly in line with neighbors’ performance, but there is a broad range of results.
Figure 3.Tax Efficiency in Balkan Countries, 2014
Sources: IMF, WEO database; and IMF staff calculations.
5. Albania’s VAT efficiency appears lower on average than other Balkan non-EU countries but above the EU performance. However, the variability of the share of domestic consumption relative to GDP biases this measure. In particular, in Albania, remittances to households by the large number of Albanian émigrés working abroad boosts consumer spending, which is the VAT base, as a proportion of total economic activity as measured by GDP. This boosts artificially the observed efficiency of VAT relative to GDP.
6. Albania’s VAT efficiency as a percent of final consumption indicates sizable revenue losses. Albania’s VAT efficiency, as measured by C-efficiency, is slightly below the European average, which is dominated by EU member states (Figure 4).3 Multiple sources have reported widespread problems with VAT fraud within the EU; these imply significantly reduced VAT collections and efficiency. In addition, most European countries have a number of material VAT exemptions and other tax expenditures, which also decrease C-efficiency ratios. Therefore, the average for Europe reflects material levels of noncompliance and tax expenditures. It follows that Albania’s C-efficiency, which is slightly below the European average, also indicates likely material losses from tax expenditures and/or noncompliance.
Figure 4.Average C-Efficiency Ratio in Selected European Countries, 2011-13
Source: IMF staff calculations.
7. The overall impact of tax expenditures in Albania is not high by European standards, except for having higher zero-income-tax thresholds. The design of Albania’s VAT follows the EU standard, with a small number of additional statutory exemptions—the main one being the exemption for sales of newspapers and magazines and media advertising. Furthermore, the Albanian VAT applies exemptions where many EU member states apply zero-rating (for example, to food). This is more expensive, because of the mechanics and timing of input tax credits.4 The personal and corporate income taxes are relatively straightforward, albeit with relatively high zero-rate thresholds. There are a number of exemptions on imports from both VAT and excises, which seem to aim at facilitating foreign investments and correcting weaknesses in the domestic VAT refund regime. The net cost of VAT exemptions in policy terms is not high as these would otherwise be claimed as input tax credit—however, they raise compliance and administrative issues, such as increased risks of misreported imports and decreased control over the supply chain for VAT, increasing the costs through lower compliance.
8. Overall, low tax revenues in Albania seem to be mostly imputable to a low level of compliance. Estimating definitively the relative impact of high thresholds and low compliance on income taxes and social contributions requires better knowledge of the distribution of individual and corporate incomes. However, officials and external observers report widespread failure to register for as well as pervasive excise smuggling and fraud. A preliminary decomposition of the observed VAT C-efficiency measure (Section III) indicates substantial VAT compliance losses.
B. Recent Developments
9. Despite several tax rate increases, tax revenues have remained broadly unchanged compared to 2009 (Figure 5). The EFF-supported program launched in 2014 helped reverse a decreasing trend in tax collections, mainly through increased social contributions and profit tax. The authorities increased tax rates across the board, in particular for corporate tax and excises. Conversely, the 2014 reform of the personal income tax—from a flat tax to a progressive system where 60 percent of the taxpayers are below the zero-rate threshold—weakened revenue collection. In the 2016 budget law, the government abrogated the small business tax and exempted small and micro businesses from any profit tax; these measures are likely to have further weakened direct taxation. The increase in excise duty rates for road fuels and tobacco also failed to increase collections in 2015, but this was mainly due to taxpayers’ pre-releasing stocks in 2014 ahead of the anticipated duty rise, thus decreasing 2015 revenues.
Figure 5.Albania: Tax Revenues
Sources: IMF, WEO database; and IMF staff calculations.
10. The VAT C-efficiency ratio in Albania has decreased from 60 percent in 2008 to 50 percent in 2013, indicating an increase in the VAT gap (Figure 6). While over 2000–2008 the VAT C-efficiency was on a rising trend, this reversal could result from an increase in the VAT gap, due to changes in either policy, behaviors, or compliance.5
Figure 6.VAT C-Efficiency Ratio
Source: IMF staff calculations.
11. The tax efficiency of CIT has also fallen, from 0.15 in 2011 to 0.10 in 2014 (Figure 7). The observed tax efficiency relative to GDP has fallen consistently throughout 2011–2014. In terms of revenue collected, this partially offset the effect of the increase of the top rate of CIT to 15 percent in 2014.
Figure 7.CIT Efficiency
Source: IMF staff calculations.
C. Revenue Management and Forecasting Errors
12. Forecasting revenue is a difficult exercise in a small open economy like Albania. The economy is subject to numerous external shocks, which are large relative to the size of the economy. For instance, foreign direct investments generally have a high import content. Hence, large foreign-financed projects routinely produce a lumpy profile of imports of goods and services and customs revenues.
13. Revenue forecasts in Albania are characterized by significant errors (Figure 8, left panel). The root-mean-square error (RMSE) of the revenue projection on which budgets are based is close to 1.5 percent of GDP. This represents a substantial forecast dispersion, which poses risks for the budget.
Figure 8.Errors in Budget Revenue Forecasts1
1 In these charts, year t refers to the year covered by each budget, and year t+1 refers to the following year. Forecasts are thus generally prepared at the end of year t-1.
14. Moreover, the forecasts have a systematic and sizable upward bias (Figure 8, right panel). On average over the last eight years, budget revenue projections have been consistently too optimistic by almost 2 percent of GDP. While such an upward forecasting bias is not uncommon, it is higher in Albania than in neighbors and peers.
D. Revenue Performance in 2015
15. In 2015, the government’s revenues missed their budget projection by 2½ percent of GDP. The 2015 forecast error exceeds the recent average. The underperformance was in almost all the main taxes, with the exception of corporate income tax and social security contributions.
16. Revenue forecasts were overly optimistic. Tax revenue forecasting used a “top down” methodology. Nominal GDP growth was applied across the board to expected collection outcomes for the base year, with adjustments for the estimated impact of policy changes and anticipated increases in collection efficiency. There was no regard for microeconomic factors particular to different tax headings and limited consideration for one-off and other distorting factors known to the revenue administrations. Jensen and others (2015) identified the following contributing factors:
The estimated 2014 base for the 2015 revenue forecast was too optimistic. Some significant base adjustment and one-off factors were not taken into account in preparing the 2015 budget, in particular the early release of excise goods in 2014, in anticipation of tax increases taking effect in early 2015.
The growth rates applied were also too optimistic. In addition, exogenous surprises contributed to the underperformance: namely, plummeting oil prices and declining deposit interest rates.
Forecast estimates for tax collection efficiency were unrealistically high. These coefficients were intended to set revenue administration targets rather than provide a realistic forecast.
Estimates for the impact of some tax policy measures were unrealistic. Behavioral responses were underestimated, such as the increased smuggling of cigarettes that contributed to a major reduction in cigarette imports.
17. Most of the underperformance of tax and customs revenues in 2015 was the consequence of overly optimistic assumptions (Table 3). A tentative decomposition of the errors shows that more than half of the overall underperformance related to unrealistic assumptions pertaining to administrative efforts and elasticities with respect to GDP—in particular, the growth of declared volumes of cigarettes and other excisable commodities. The remaining underperformance can be attributed to the decline in oil prices (26%) and the shortfall in nominal GDP growth (13 percent).
|2015 performance||Breakdown of the deviation|
|(in ALL billion)||Budget||Est.||Deviations||Base|
|Social security contributions||68.8||72.9||4.1||6.7||−1.1||−1.0||−0.3||−0.2|
E. Compliance and Evasion
18. There is a consensus that Albania suffers from widespread informality, with a significant adverse impact on tax revenues. Government officials and external observers consistently portray widespread noncompliance across most taxes. Further evidence of endemic compliance issues is provided in the performance data compiled by the customs and tax administrations, which show a high percentage of compliance checks yielding additional liability (up to 100 percent in some regions). Whilst such high hit rates may be partly the result of better, risk-based case selection, they also imply that a large number of noncompliant taxpayers remain to be identified and assessed for additional liabilities.
19. High levels of noncompliance not only damage tax revenues, but can also undermine social cohesiveness and economic growth. There has been a number of empirical studies establishing causal links between tax compliance and social cohesion. In particular, taxpayers need to perceive tax policy as fair, effectively administered (so that everyone pays their fair share). There must be a clear, transparent link between tax payments and public benefits. Enterprises thrive in countries with properly funded public services and infrastructure, where revenue administrations facilitate compliance and unfair competition from noncompliant businesses is minimized.
20. While tax gaps in Albania have never been quantified, an approximate level of VAT noncompliance can be derived from C-efficiency. C-efficiency measures the efficiency of the VAT tax base (i.e., final consumption) in generating VAT collections (section I). Efficiency losses can stem from structural reliefs and tax expenditures (the policy gap) and noncompliance (the compliance gap).6 The approximate compliance gap can therefore be derived from C-efficiency and the likely policy gap.
21. The compliance gap for VAT is provisionally estimated to be in the region of 35 percent of potential VAT. Previous research has found that virtually all EU member states have policy gaps in the range 20–35 percent (EC, 2013). The design of the Albanian VAT is much the same as the EU standard, with limited additional tax expenditures. So, Albania’s policy gap likely is in the range 25–30 percent. Given Albania’s observed C-efficiency of 46 percent in 2013, we estimate the compliance gap to be around 34–39 percent of potential VAT.
22. Weakening C-efficiency over recent years may be due to increasing VAT noncompliance. C-efficiency in Albania has fallen steadily over 2008–2013. Assuming that the composition of final consumption remained largely unchanged over that period, with no substantial increases in the policy gap due to new tax expenditures, the implication is that the declining trend has been caused by worsening compliance.
23. The compliance gap in excise duties is likely significant too. Officials and external observers believe that there is substantial smuggling and fraud in excise duties in Albania, particularly in road fuels and cigarettes, based on evidence from compliance and enforcement operations. As part of the recently launched campaign against informality, measures have been implemented to counter such noncompliance, including a review of the excise stamps regime and strengthened control of trans-shipments of excise goods across Albania, to prevent diversion to untaxed domestic markets.
F. Policy Recommendations
24. Low tax revenues stem from weaknesses in tax administration and compliance, which are being addressed. The Albanian authorities embarked on a reform effort in September 2015 to improve tax compliance, fight evasion, and minimize informality—and so increase tax collections. The effort is multifaceted. The government launched a public awareness campaign, waived penalties for businesses that become fully compliant before end-2015, implemented pilot compliance campaigns based on centralized risk profiles and set up a lottery to incentivize customers to claim their tax receipts. To enhance enforcement, 500 new tax inspectors were hired. A steering committee monitors the effort on a weekly basis, and promotes closer collaboration between tax and customs administrations.
25. The compliance campaign should focus more on long-term outcomes, using expertise and new risk profiles from the joint tax and customs risk unit. The risk analysis from the newly merged tax/customs risk unit and operational intelligence has informed the compliance measures implemented in the informality campaign. These measures are having an impact in terms of increasing taxpayer registrations and early indications of improved revenue yield from compliance action. For example, verification of 1,358 VAT taxpayers by March 8, 2016, yielded an additional ALL 2,961 million, with a hit rate of 81 percent. Individual regions’ results varied widely in terms of both take up of the initiative and hit rate, but the Tirana region recorded a 100 percent hit rate on 651 cases with a yield of ALL 1,585 million and the large business office recorded ALL 1,141 million from just 8 cases, out of a total of 15 checked. There is clearly a strong direction behind the campaign, and the authorities need to continue refocusing compliance interventions on high value, high risk taxpayer segments and on implementing modern compliance risk management frameworks.
26. The informality campaign should establish a robust performance-monitoring regime focusing on a relatively small number of key strategic performance indicators. While detailed reporting of inputs, actions, and outputs is critical for operational managers, there is a need at the strategic level to monitor the overall outcomes of the informality campaign and maintain clear visibility of strategic objectives.
27. In terms of tax policy, the authorities should seek to simplify the tax system, so as to reduce administrative and compliance costs. Surveys show that such costs are relatively high (Figure 9). Implementation of a mandatory online filing of sales transactions ledgers should be viewed with caution given potential compliance costs for taxpayers, capacity risks for the tax administration and financial costs. Eliminating import tax exemptions (see ¶7) would also reduce the risks of misreporting and facilitate proper monitoring of the supply chain for VAT system. Relatively high zero-rate thresholds make it more difficult for the administration to establish taxpayers’ liability to register and pay tax—because it is more plausible for taxpayers to claim that their income is not high enough to bring them into tax.
Figure 9.Administrative Efficiency in Paying Taxes
1/ Each unit in the chart corresponds to 10 hours.
Sources: Doing Business (2016); and IMF staff calculations.
28. The authorities should continue to review the campaign’s governance, establishing clear and transparent lines of responsibility. High-level support was vital for the launch and early stages of the campaign. As the campaign progresses, its day-to-day governance can be delegated to senior managers in the revenue administrations and the Ministry of Finance, with continuing oversight by the Minister of Finance. Operational decisions should be insulated from political considerations and the campaign’s governance become part of the new corporate strategy for tax administration.
29. Increased cooperation between customs and tax administrations in joint risk analysis and compliance operations is desirable. The merger of the tax and customs administrations remains a long-term goal, but is not currently being pursued, to avoid distraction from the informality campaign. Nevertheless, there have been initiatives to increase cooperation and joint working between the two agencies. These include the establishment of a joint risk unit, increased data sharing and pilot programs for joint audits. These initiatives are to be encouraged and maintained.
Alm, J.,2013, “Expanding the Theory of Tax Compliance from Individual to Group Motivations,” Working Papers 1309 (Tulane University).
Alm, J. and B.Torgler,2011, “Do Ethics Matter? Tax Compliance and Morality,” Journal of Business Ethics, Springer, Vol. 101(4), pp. 635–51.
Alm, J. and B.Torgler,2005, “Culture Differences and Tax Morale in the United States and in Europe,” Journal of Economic Psychology, Vol. 27(2), pp. 224–46.
Alm, J., K.M.Bloomquist, and M.McKee,2013, “When You Know Your Neighbor Pays Taxes: Information, Peer Effects, and Tax Compliance,” Working Papers 13–22 (Appalachian State University).
Bazart, C. and M.Pickhardt,2009, “Fighting Income Tax Evasion with Positive Rewards: Experimental Evidence,” Working Papers 09-01, LAMETA (University of Montpellier).
Barretto, R.A. and J.Alm,2003, “Corruption, Optimal Taxation, and Growth,” Public Finance Review, Vol. 31.
Calvet, R. and J.Alm,2014, “Empathy, sympathy, and tax compliance,” Journal of Economic Psychology, Vol. 40(C), pp. 62–82.
Daude, C., H.Gutiérrez, and Á.Melguizo,2012, “What Drives Tax Morale”, OECD Development Centre, Working Paper 315.
European Commission, 2010, “Compliance Risk Management Guide for Tax Administrations,” TAXUD.
European Commission, 2013, “Study to Quantify and Analyse the VAT Gap in EU Member States,” TAXUD/2013/DE/321.
Feld, L.P., B.S.Frey, and B.Torgler,2006, “Rewarding Honest Taxpayers? Evidence on the Impact of Rewards from Field Experiments,” CREMA Working Paper Series.
Jensen, A., S.Caner, E.Hutton, M.O’Grady, and S.Hansen,2015, “Causes of Tax Revenue Underperformance and Options for Corrective Measures,” IMF Technical Assistance Report.
Joshi, A., W.Prichard, and C.Heady,2014, “Taxing the Informal Economy: The Current State of Knowledge and Agendas for Future Research,” The Journal of Development Studies, Vol. 50(10), pp. 1325–47.
Kanbur, R. and M.Keen,2014, “Thresholds, informality, and partitions of compliance,” International Tax and Public Finance, Vol. 21(4), pp. 536–59.
Kastlunger, B., S.Muehlbacher, E.Kirchler, and L.Mittone,2011, “What Goes Around Comes Around? Experimental Evidence of the Effect of Rewards on Tax Compliance,” Public Finance Review, Vol. 39, pp. 150–167.
Keen, M.,2008, “VAT, tariffs, and withholding: Border taxes and informality in developing countries,” Journal of Public Economics, Vol. 92(10–11), pp. 1892–1906.
Keen, M.,2013, “The Anatomy of the VAT,” IMF Working Paper 13/111.
La Porta, R. and A.Shleifer,2014, “Informality and Development,” Journal of Economic Perspectives, American Economic Association, Vol. 28(3), pp. 109–26.
Luttmer, E.F.P. and M.Singhal,2014, “Tax Morale,” Journal of Economic Perspectives, American Economic Association, Vol. 28(4), pp. 149–68
OECD, 2014, “Compliance Risk Management: Managing and Improving Tax Compliance,” Guidance Note.
Saez, E.,2010, “Do Taxpayers Bunch at Kink Points?” American Economic Journal: Economic Policy, 2(3), pp. 180–212.
Prepared by Nicolas End and Mick Thackray.
This measure provides a broad indicator of the efficiency of each tax as being the amount of additional revenue produced per percentage point of the top rate. Its value is a function of (i) the coverage of the tax base by tax policy and degree of progressivity, (ii) the composition of GDP and distribution of incomes/expenditure; and (iii) the level of taxpayer compliance.
Formally, C-efficiency is computed as follows:
Especially when refunds are restricted.
An alternative explanation is a change in the composition of GDP in Albania, for example increasing consumption of exempted commodities. This possibility will be covered by the upcoming RA-GAP study of the VAT gap.
Keen (2013) formally decomposes C-efficiency as: C-efficiency = (1 − policy gap) × (1 − compliance gap).