Central Eastern Europe and the “Flat” Tax - a follow up

The issue of flat taxes in CEECs has been given wide attention, including posts on this blog (see Taxes – linear or not?), though (as we mentioned in the previously cited post) for a number of reasons the arguments in favor of “flat” taxes in the CEECs are rather anecdotal than backed with theoretical analysis or empirical facts.
A September IMF working paper (IMF WP/06/218) by Keen, Kim and Varsano attempts to fill part of this gap, by looking at the principles and evidence on the rationale behind flat taxes (admittedly putting more weight on the principles than the evidence).

They start of from (justly) underlining the key issue of the "flat" taxes in the CEECs - as we claimed in the blog, they are actually rather progressive than flat. The presence of any sort of tax-free threshold (TFT) causes a tax system with a linear tax of A% above that level, to be in fact nothing different from a progressive tax scheme with two income groups: (1) income TFT where income-TFT is taxed at a A% rate. This complicates the deduction at source (as individuals may receive income from more than one source, See box 1 here) and therefore the largest benefits of a pure linear tax cannot be reaped. In general the only country (from our sample) with what can be labelled a pure flat tax is Georgia, while Russia and Ukraine being relatively close to such a system for two reasons - one the TFT is rather low, and second, it disappears for individuals who exceed a certain income - facilitating at least some types of deduction at source.
An interesting point is mentioned about social contributions is that social secutrity contributions and their 'loose' link to future social security payments may (and in practice do) seriously distort the 'flatness' of taxes.
Consequently we will use the nomenclature of the paper and distinguish a ‘pure flat tax’ from a ‘flat tax’ – the latter being a more broad definition encompassing the systems in the sample (see table 1).

Next, the authors draw a distinction between what they call two waves of flat tax introductions – the first wave, pursued by the Baltics in the 1990s tended to set tax rates at rather high levels, close to the highest rates of the previous progressive tax while in the later wave, starting from Russia, PIT rates were set close to the lowest pre-reform tax rates, thereby leading to a reduction of the marginal tax rate for most income groups.
One thing they seem to omit is the fact that these two groups are in a sense converging – PIT rates are being reduced in Estonia and Latvia, while being increased in Ukraine. Further the countries differ on the approach to CIT, dividend tax, capital income tax, VAT and TFT (except for Latvia and Georgia the TFT was raised during the tax reform).
Aside the detailed review of the reforms, the authors focus on analyzing the principles of flat taxes. They discuss whether a flat tax can be ‘optimal’ from the point of view of balancing efficiency costs to incentive distortions against the benefits of some redistribution (and find that it may not, but also that nothing guarantees that the progressive systems in place are), whether it reduces the incentives for tax avoidance and tax evasion (and though it seems it should, in principle may also cause the contrary), whether it is politically appealing (and what chances are the switch to flat taxes will be soon reversed). Next they look in detail at the possible effect (in principle and, admittedly slightly less convincingly, in evidence) on the distributional effect between income groups, on work incentives (the changes affect both the marginal and average tax rates and thus results in an ambiguous outcome of the interplay of the income and substitution effects), compliance, ease of administering, simplicity (all of which, surprisingly even the last one, according to a Russian-taxpayer survey seem ambiguous) and automatic stabilizers(where the flat tax with TFT seems to have some potential).

Generally the paper is a comprehensive study of the potential rationale behind flat taxes. It provides arguments why the benefits of flat taxes are not so clear in the first place, and why it is so difficult to analyze the effect of flat taxes that formed part of comprehensive tax reform packages (it is truly challenging to disentangle and identify the individual effects of the reforms in the packages - which were not only often not explicitly intended as revenue neutral, but consisted of: the introduction of a flat tax together with a change in tax rate(s); a change in TFT; widening of the tax base via the reduction of exemptions; general attempts to simplify the tax calculation and payment procedures; reforms in tax collection, monitoring, penalty and enforcement. In many cases this may prove unconvincing if not impossible).
This makes the literature review a very interesting and welcome reading.

To me, however, the “flat tax” remains more of a marketing slogan, used to publicize tax reforms – and to signal a change in the regime and the approach to taxes and market economy in general. This may not always prove a successful mean as it tends to spur the backlash of arguments about its alleged non-progressiveness, and thus the breach of the social contract. This, together with a different (at least perceived) effectiveness of the pre-reform or existing tax systems is probably the reason why the appeal was rather strong in CEECs and rather weak in developed countries. I accept the rationale behind the pure flat tax, but until I see convincing arguments in favour of the flat tax, I am somewhat reluctant to believe why it should be significantly better than any other tax system simplification.

One thing the paper lacks perhaps is some focus on the weight of the income groups i.e. when discussing the potential gainers and looser, it would be nice to see, how many people would fall in each threshold, and also what was the use of tax exemptions. Additionally some rough empirical facts, on how the declared taxable income figures changed, and consequently how the distribution changed once the reform came into life, which would yield some information on compliance – I do not know how available to IMF economists such data can be, but perhaps its worth a try. Admittedly both of these issues could well be material for another paper.

Among minor issues that arise from this report are some corrections that we have made to our table in the Taxes – linear or not? post . Firstly we have inserted the previously omitted post-reform CIT rate in Romania (which is 16%), and noted the existence of the additional total income taxin Serbia, which further strengthened in 2006 and causes the label of a “flat” tax rate an inadequate description of the
Serbian tax system. Thanks to the IMF report we were able to spot these issues and correct them. However the report itself seems to have a problem with the Estonian rates – in 2006 the PIT rate in Estonia is 23% as is the rate (not 24% as the report claims) and both will be reducing by 1% point a year to reach 20% in 2009. Although the authors focus on the rates before and after the initial reform, they try to provide up-to-date information on tax rate changes - however some issues seem to go unnoticed – the reduction of Lithuania’s PIT rate from 33% to 27% on the 1 of July 2006 and its further planned reduction in January 2008 to 24% (according to the Lithuanian Ministry of Finance ), Russia’s decrease of the CIT rate to 24% in 2002 and Ukraine’s increase of the PIT rate from 13% to 15% which is due January 2007.
Most of these have been already the table under the “old” post, but we repost a more legible excerpt from the table here:
Table 1. Tax rates in “flat tax” CEECs (updated table) as of 2006.
Sources: see (primary table), IMF(2006),
Notes: (a) 0% rate on retained earnings, paid out earnings tax at 23%, decreasing 1%point a year to 20% in 2009; (b) decreasing 1% point a year to reach 20% in 2009; (c) decreased to 27% on 01/07/2006, will decrease to 24% on 01/01/2008; (d) Serbia has an additional total income tax for income above a threshold (which decreased in 2006) thus is basically not a “flat tax” regime; (e) will increase to 15% 1/01/2007; (f) standard rate, (g)the TFT only exists up to a certain level of income.