Tony Levitas, Brown University, SKL International, The Decentralization Support Project

1. Introduction[1]

On April 29^th, 2021, the Office of the President issued a decree giving the Cabinet of Ministers three months to submit to the Verkhovna Rada legislation that would “credit part of the personal income tax to local budgets according to the registered place of residence of a natural person - a taxpayer.” The decree came in the middle of an intensifying parliamentary debate about how Personal Income Tax (PIT) should be shared with local governments, a debate that has so far generated at least 11 different proposals[2]. Of these, four call for PIT to be shared with local governments in accordance with where people work; one proposes that PIT should go to the local governments in which taxpayers live; two others suggest that PIT should be split –in different proportions-- between the local governments in which people may live and work; and two others support giving taxpayers who live and work in different local governments the right to choose where their PIT shares should go.

The attention that the Office of the President and the Verkovna Rada are paying to the “PIT Question” is extremely welcome because the current system for allocating the tax to local governments is deeply flawed and must be changed. To make a long story short, under the existing rules some local governments –particularly Kyiv and other big cities—now receive very significant amounts of PIT from taxpayers who neither live nor work on their territories, and for whom they provide no services. Meanwhile the local governments in which these taxpayers actually work and may well live, get nothing. Systemically, this makes no sense. It is also obviously unfair and --less obviously-- inefficient because it requires that the equalization system be larger than it need be.

But if there is a growing consensus that the existing system must be changed, there is also much contention and confusion about how it should be fixed. In the following, we argue that the answer to the question about how the system should be fixed is both simpler and more complicated than of any of the current proposals suggest. It is simpler because the way Ukraine should share PIT with local governments is straightforward: All PIT shares should be returned to the local governments in which taxpayers live and vote. Ukraine should do this for the same reasons that it is done in virtually all other European countries in which the personal income tax is a significant source of local government revenue: To increase the responsiveness and accountability of local governments by aligning the economic interests of citizens as taxpayers with their political interests in democracy as voters.

And it is more complicated than any of the current proposals suggest because even with the support of the Presidential decree, achieving this objective –or something close to it-- will require 18 to 24 months of analytic, technical, and political preparation. That said, we think that with the necessary investment and political will, it should be possible for Ukraine to allocate PIT to local governments by place of residence by fiscal year 2023, so long as place of residence is understood to mean where people actually reside, and not where they are currently legally registered to live.

The paper is organized as follows. Section I reviews the history of the current situation and explains why the system for allocating PIT must be reformed if Ukraine’s intergovernmental finance system –and with it, the country’s hard-won decentralization effort—are to be put on a stable fiscal foundation. Section II explains why virtually all European countries that use PIT to finance local governments identify the origin of the tax as the residency of the taxpayer, and not his or her place of employment. And Section III outlines a strategy for quickly moving Ukraine to a system in which the national government takes over from firms the responsibility for allocating PIT payments to local governments and begins to do this according to where people actually live –though not necessarily where they are legally registered to reside.

2. The History of the Current Situation in Ukraine

Before discussing why Europe shares PIT in accordance with a taxpayer’s residency, or what Ukraine needs to do to join the club, it is important to review how PIT is currently being shared with local governments. This is necessary because until recently the popular understanding has been that PIT is shared with local governments in accordance with a taxpayer’s place of employment.

In short, it has never been true in Ukraine that PIT is being allocated to local governments according to where people work.  Or at least, it has never been true for most employees of large multiunit companies who work in local governments other than the one in which their employer is legally registered. Instead, “loopholes” in the tax code have allowed firms with multiple business units to credit the PIT payments of all their employees to the local governments in which they are legally registered, and not to the local governments in which their employees may actually work. [3]

These “loopholes” are convenient for businesses because multi-unit firms do not have to trouble themselves with sending the PIT contributions of their employees to the tax authorities of all the jurisdictions in which their employees may work. Indeed, large employers have actively resisted reforms that would require them to send the PIT contributions of their employees to the local governments in which they really work. This is not surprising because such requirements would force many firms to change their accounting and IT systems, changes that obviously come with unwanted costs.

More importantly, firms should not be in the business of allocating PIT at all. Instead, this function should be –as it is elsewhere in Europe-- the responsibility of the national government.[4] To be sure, firms need to send the PIT payments of all their employees to the tax authorities, and these payments must contain the tax identification numbers of all their employees and wages received. But it is the national Treasury System that should be responsible for crediting PIT shares both to the national budget, as well as to the budgets of the appropriate oblasts and hromada.[5]

The loopholes in the law which allow multi-unit businesses to credit the PIT payments of all their employees to the local government in which they are legally registered have also benefited big cities. This is because a disproportionate number of large, multi-unit firms have their legal headquarters in big cities and thus often receive the PIT shares of people who work (and live) elsewhere --and for whom they provide no services. Meanwhile, and conversely, the smaller, generally poorer local governments in which these taxpayers actually work --and typically live-- must provide them with services but do not get their PIT payments. This makes no sense and is obviously unfair. It is also inefficient: Sending the PIT payments of people who live and work in smaller, poorer hromada to the bigger, wealthier cities in which their employers are headquartered increases the need for equalization and requires that the base/reverse grant system be larger than necessary.

But if the existing system has been convenient for big business and profitable for big cities, it is nonetheless the Ukrainian state that bears the greatest responsibility for failing to fix it. For years, it has been open secret within the Ministry of Finance that significant amounts of PIT are being sent to local governments in which people do not work -or for that matter live- and for whom these local governments provide no services. Nonetheless, the Ministry has not pushed to reform the existing system. And it is easy to understand why: The Ministry has had its hands full dealing with an already daunting list of structural reforms in a period of great socio-economic turbulence, to say nothing of war. Moreover, the Ministry knows that any serious reform of the allocation of PIT will require major changes in the way the Tax Administration in particular, and the Ukrainian state in general registers, links, and makes use of the data on where its citizens live, work, pay taxes, and vote. Indeed, the Ministry is painfully aware that meaningful reform will not only require coordinating the work of multiple ministries but developing and implementing the information and data processing systems necessary for the State Treasury –and not firms—to allocate PIT to local governments.

All this is true. But the central argument of this paper is that the decentralization reforms that the Ukrainian state has so heroically pursued over the last few years simply cannot be put on a sound foundation without sharing PIT with local governments in accordance with where taxpayers live. And that this cannot be done without first linking people’s PIT payments to their place of residence and then, by making the Ukrainian state responsible for allocating these PIT payments to the local governments in which they actually reside.

In the concluding section of the paper, we will outline how this might be done. But for the moment, the critical thing that needs to be understood is that the national government cannot allocate PIT to local governments in accordance with where people work because it does not know where people work. And without knowing where people work, it cannot link data on citizens PIT payments with data on the local governments in which they are employed. Indeed, it is precisely because of this, that the national government is relying on private companies to divvy up PIT payments between itself and local governments.

Worse, and the reason this is a nontrivial problem, is precisely because we know –though we cannot say exactly how much-- that very significant amounts of PIT are being sent by firms to the local governments in which they are headquartered, and not to the local governments in which their employees work. For example, we know that 14% of all firms are legally registered in Kyiv and that they account for 24% of national employment. But only a small fraction of that employment take place in the capital.[6] Anecdotally, we also know that both many large state enterprises and private firms register the PIT of all their employees in the local government in which they are headquartered –be that in Kyiv or some other city.[7]

But the most compelling indication of the magnitude of PIT payments that are being misallocated under the current rules comes from Kyiv’s budget: In 2020, Kyiv’s per capita PIT revenues were more than three times higher (8,400 hr./pc) than those of the average hromada (2,700 hr./pc).[8] This 3 to 1 ratio is on the high end of the spectrum but in and of itself, is not particularly unusual: Capital cities often have per capita tax revenues two or three times those of the average local government –and often 6 or 7 times more than the poorest— both because they are the hubs of the national economy and because wages in them tend to be substantially higher than elsewhere.

But what is unusual, is that Kyiv is achieving this 3 to 1 ratio with a 40% share of PIT when all other hromada receive a 60% share. In other words, the yield of Kyiv’s PIT share relative to the average hromada would be more or less understandable if all local governments received the same share. But it is exceptionally high --and hard to explain-- when we consider that Kyiv gets one third less of every taxpayer’s PIT payment than all other hromada[9]. In short, Kyiv is making up for its smaller share through the PIT payments of taxpayers who do not work (or live) in Kyiv, meaning the employees of large, multi-unit enterprises who work --and almost certainly live-- outside of the capital.

Looking forward, the most important implication of all this is that if PIT really were to be allocated by place of work, Kyiv would lose the revenues it currently receives from people working outside of the capital but employed by firms registered in it. Or put another way, if PIT was actually allocated by place of work, Kyiv’s PIT share would have to be significantly increased for the city to maintain anything like its current revenues from the tax. We will return to this point later. But for the moment, the point worth holding on to is that Kyiv PIT share will have to be renegotiated upward whether the origin of PIT is defined as place of place of work or place of residence because its current 40% share only works financially because of the PIT payments the city is receiving from people who neither work nor live in the capital.

In short, and to summarize the discussion so far, PIT is not now being allocated to local governments by place of work, though for years this is what people have believed. Instead, very significant amounts of PIT are being misallocated to big cities because firms --not the state-- are responsible for sending PIT payments to the appropriate tax authorities, and because firms with multiple business find it easier to credit the PIT of all their employees to the jurisdiction in which they are legally registered, and not to the many different local governments in which their employees may work. This is unfortunate, but firms should not be in the business of divvy-up public revenues.

Moreover, the misallocation of funds is not trivial if Kyiv’s per capita revenues from PIT are 3 times the national average despite receiving a PIT share 30% lower than all other hromada.[10][11].  Worse, the misallocation of these revenues to Kyiv –no matter how big or small—underestimates the misallocation across all hromada-- because there are many multi-unit businesses who are headquartered in local governments other than the capital. More importantly, and looking forward, while the national government knows that a lot of PIT is being sent to the wrong local governments, it cannot say by how much or to whom because the Tax Administration cannot technically link people PIT payments to their place of work (or residence). Finally, and most importantly, allocating large amount of peoples’ PIT payment to local government in which they neither live nor work is unfair, inefficient, and makes no systemic sense.

3. Why the Answer to the PIT Question Should be Straightforward

Last year, Ukraine completed a heroic effort to consolidate over 10,000 tiny villages into hromada capable of providing their residents with basic public services. Many outstanding issues, including constitutional ones, remain to be resolved before policy makers will be able to say that Ukraine’s subnational finance and governance system rests on a stable foundation. PIT, which accounts for more than 50% of all hromada revenue (including Kyiv), and about 70% of all oblast revenue is obviously one of the cornerstones of this foundation. And until the allocation of PIT is put on a sound basis, the foundation of local government finance and governance in Ukraine will by shaky.

The constitution of new hromada, and the fact that many of their elected officials have come to understand that they are not receiving the PIT payments of people who work (and/or live) on their territories has increased the awareness of both parliament and the President’s Office that the existing system for sharing PIT needs to be changed. The question is how. Here, it is worth looking first at a little public finance theory, and then more extensively at the practices of other European countries that rely heavily on PIT to finance subnational governments.

Perhaps the most important normative claim of the literature on subnational finance is that local governments should --to the greatest degree possible-- pay for what they do from taxes, fees, and charges whose rates they set for themselves[12]. The argument behind this claim is that the more a local government’s revenues are determined by democratically elected councils, the more likely it is that these councils will make tax and spend decisions that are in line with their electorate’s preferences: If councils tax too much, or fail to deliver services that people want because they tax too little, the people will –as the saying goes-- “vote the bums out office.” In short, much of the literature on public finance insists that local taxation is critical for ensuring that local governments are responsive and accountable to their electorates.  

Much can be said about this central pillar of the literature on subnational finance. But two points are worth stressing. The first is simply that in many, maybe even most European countries the largest source of local government revenue does not comes from taxes that local governments set themselves. Indeed, there are relatively few countries in which the normative principle of the literature is really respected in practice, or at least respected to a degree that would satisfy its most ardent proponents.[13]

The reason for this contradiction between theory and practice has two sides and is easy to understand once both of them are put together. On the one hand, the highest yielding taxes of the contemporary nation state –the Value Added Tax, the Personal Income Tax, and the Corporate Income Tax— are all more effectively and efficiently administered by the national government than by local ones.  On the other hand, the cost of the public services that are often assigned to local governments –particularly social sector functions like education, housing, health, and community services-- far exceed what can possibly be raised through what the literature considers to be good local taxes, meaning locally set user charges and fees, and above all the property tax.[14] In short, in many countries in which local governments perform significant social sector functions, the lion’s share of their revenues come not from local taxes but from some combination of categorical grants, freely disposable general revenue grants, and shares of national taxes.

But the second thing that should be said about the literature’s admittedly sometimes obsessive concern with local taxation, is that even if local governments are typically heavily dependent on grants and shared taxes, it still makes sense for citizens to pay local taxes, and to know that they have a financial stake in the behavior of their elected officials. In short, people are more likely to pay attention to what their local governments do if they know they are being taxed to pay for at least some of the services they receive. Indeed, the basic gist of this argument can be extended to shared taxes: Even if local councils do not, for example, set PIT rates, local voters are much more likely to be concerned with how ‘their’ local governments spend money if they know that much of that money comes from ‘their’ PIT payments and not from general budget of the national government.

With these points in mind, now let us look how PIT is used to finance local governments in Europe.  In at least twenty European countries, PIT revenues constitute more than 20% of local budgets. Moreover, in many countries PIT is often the largest single source of local revenue.  But there are three distinct ways PIT can be used to finance local governments: By earmarking a percentage of the national yield of PIT for local governments and then allocating the resulting pool of funds to them by formula; by giving local governments some control over PIT rates, while base setting and collection remain in the hands of the national government; and by giving local governments a fixed share of the PIT generated on their territories.

Both the Czech and Slovakia Republics earmark a percentage of the national yield of PIT for local governments and then allocate these monies to them by formula. Here, there is no relationship between the amount of PIT that individual local governments receive from the tax, and the amount of PIT paid by the people who live or work on their territories. Indeed, the national government does not need to know where PIT comes and has no need to link PIT payments with either peoples’ place of work or residence. Instead, the ‘PIT share’ functions as a legal ‘anchor’ for a pool of national funds that are then divided back down to local governments by formula through freely disposable general revenue grants. This ‘anchor’ or ‘peg’ also serves to ensure that the overall grant pool created by earmarking a share of the national yield of PIT for local governments grows with the economy as a whole.

The Czech and Slovak intergovernmental finance systems are thus conceptually closer to systems that rely on general revenue grants to finance local governments than to systems that provide local governments with large chunks of revenue by allowing them to retain some of the PIT generated on their territories, either through rate setting powers or a guaranteed fixed share. The Czech and Slovak systems have the virtue of being relatively simple to administer. They are also strongly equalizing because the most important coefficient used in allocating general revenue grants is almost always population. As a result, most local governments get the same basic payment per inhabitant, making other efforts to compensate local governments for differences in their revenues or service costs less important systemically. But these sorts of systems divorce local government revenues from both the pocketbooks of voters and the health of the local economy: The former reduces the incentives of citizens to oversee and hold their local officials accountable, while the latter reduces the incentives of local officials to grow the local economy.   

This stands in stark contrast with systems that give local governments either a share of the PIT generated on their territories and/or the right to set PIT rates. Now, differences in the relative wealth of local governments --in their PIT bases-- matter a lot. Local governments with high unemployment, low wages, or low official employment, will get relatively little out of the tax, either through shares or rate setting powers. Afterall, the right to tax a weak base, or to receive a hearty share of almost nothing, will yield little in budget revenue.  

Intergovernmental finance systems that provide local governments with large amounts of revenue from the PIT generated on their territories therefore require more visible equalization mechanisms. And there are many ways that equalization can be done. But one simple mechanism often used in countries that share PIT with local governments is to give poorer ones equalization grants that bring their per capita PIT yields up to some reasonable percentage of the national average –usually 75-85%. It is also common for richer local governments to contribute to some of the costs of equalization by ‘clawing back’ some of their ‘excess’ PIT revenues. This is what Poland, Germany and Ukraine all do, though in Ukraine the richest local government –Kyiv—does not contribute to the equalization system, a point which we return to later.

Things get a little more complicated in countries in which local governments have been given some rate setting powers. Nonetheless, the same basic idea is often applied: Here, the national government must know the PIT rates set by all local governments so that it can determine a national average rate. The average rate is then applied to the PIT bases of all local government to determine both the per capita PIT baseline for equalization, and the per capita PIT yield of every local government at the standardized rate. This procedure prevents poorer local governments from seeing their equalization payments fall if they chose to tax PIT at a higher-than average rate. And it prevents richer local governments from gaming the system to get equalization grants by choosing to tax at a lower-than average rate.  

All of which is to say that when local governments get significant amounts of revenue through the PIT generated on their territories, equalization becomes more important and more burdensome, particularly if local governments are given rate setting powers. Such systems also require the national government to link PIT payments with taxpayers’ place of residency (or work) in order to allocate PIT shares to the appropriate local government. As a result, origin-based PIT sharing systems are more complicated to administer than their Czech or Slovak counterparts. But unlike these counterparts, or more generally, systems that rely on providing large sums of money to local governments through general revenue grants, systems that rely on origin-based PIT sharing link the financial interests of their electorates with the performance of their local governments, a linkage that serves to strengthen political accountability.  They also link local government revenues with the health of the economies in which they operate, a linkage that incentivizes local officials to promote job growth and development.

Now, let us look a little more closely at the European countries that have given local governments some control over PIT rates, and then at those that share PIT with local governments on an origin basis. At least nine European countries --Sweden, Norway, Denmark, Iceland, Finland, Belgium, Switzerland, Italy, and Croatia— have transformed PIT, or at least part of it, into a true local tax. And at least another ten share a significant percentage of PIT with local governments on an origin basis. This group includes Germany, Poland, Latvia, Estonia, Lithuania, Serbia, Bosnia and Herzegovina, and Romania.  

The countries that have given local governments rate setting powers over PIT have done this by allowing them to set rates within upper (and usually lower) limits established by the national government, while also leaving the national government fully responsible for defining the base of the tax and for collecting it. The degree of local control over PIT rates varies widely across the group. At one end of the spectrum are Sweden and Denmark – arguably the two most decentralized countries in the world. Here, locally set PIT-rates generate about 90% of the tax’s total yield, meaning the national government gets almost nothing from PIT. More commonly, “rate space” is divided between the national government and local governments so that both levels of government have room to vary their rates, and both levels derive significant revenue from the tax. This is what is done in Norway, Finland, and Iceland for local governments; for regional governments in Belgium and Italy; and for both in Switzerland.

Meanwhile, Croatia sits at that the other end of the spectrum by combining very limited local government rate setting with PIT sharing: Local governments get most of their revenue from a 40% share of PIT.  But big cities can impose a local surcharge of up to 2% on the national government’s PIT rate. This arrangement preserves national control over rate setting. But it also enhances the accountability of big cities to their electorates while also giving them some independent revenue raising powers. These are attractive systemic features that are worth considering in Ukraine.

But more important for our purposes, all countries that give local governments rate setting power overs PIT define the origin of PIT as the taxpayer’s principal residency, not his or her place of work. There are three basic reasons for this. First, the basic idea of PIT is that it is a tax on peoples’ total income, not just on their wages. One of the objectives of the tax, then, is to aggregate –particularly at the upper end of the income scale-- income that can come from multiple forms of employment, as well as from rents, dividends, and interest. These sorts of income are often taxed separately or little at all. But countries that adopt PIT typically try to bundle them into the tax over time, though speed with which this is done, and the types of incomes that are eventually bundled into PIT, vary greatly. Nonetheless, because each of these incomes can come from different places it makes no sense to consider PIT as having any other point of physical origin than the taxpayer’s principal place of residency. In this sense, attributing PIT –as opposed to a wage tax—to a person’s place of employment conflicts with the basic nature of the tax.

The second, and more important reason that the origin of PIT is defined as a taxpayer’s residence in all countries that give local governments rate setting powers is to ensure that political and fiscal jurisdictions are the same. This is necessary because it makes no sense to allow local governments to impose taxes on people who do not vote for them, as would be the case if local governments could tax the incomes of people who worked in their jurisdictions but did not live or vote for them. Indeed, this would be a good example of “taxation without representation”.  

Equally importantly, the logic of unifying fiscal and political jurisdictions holds true even if PIT is only being shared with local governments. After all, what sense would it make to say to citizens that 60% of your income is going to support a local government, but maybe not the one you voted for or live in. In short, as soon as we say that individual local governments can either set PIT rates or get a fixed share of the tax generated on their territories, then the only reasonable way to define the origin of PIT is by the residency of the voter-taxpayer. Otherwise, we would be allowing local governments to spend the money of people who do not vote for them and to whom they are not politically accountable.

Finally, the third reason that the origin of PIT is defined as a taxpayer’s residence in almost all European countries is that the essential purpose of local governments is to provide services to their residents. This, of course, does not mean that local governments do not provide some services to non-residents. They do. Everywhere in the world, people commute to work and make use of services provided by the local governments in which they are employed. Moreover, big cities frequently provide services to people who neither live nor work in them, but who visit to enjoy amenities, to purchase commodities, or to access national or regional services that are only provided in larger jurisdictions.

So, the question is not whether some local governments pay for the services of non-residents, or even that these costs cannot be quite significant. They often are, particularly for capital cities and other major urban centers. Indeed, countries frequently take measures to compensate big cities for these additional costs. For example, when countries rely primarily on general revenue grants to finance local governments --as is the case of Czech and Slovakia-- these grants often have ‘U-shaped’ distributions which provide more per capita revenue to both sparsely and densely populated jurisdictions. These grant systems, in other words, attempt to compensate sparsely populated local governments for the additional costs of serving dispersed populations, while also providing more per capita revenue to densely populated ones to compensate them for some combination of the costs of serving significant numbers of non-residents and the costs of providing services in areas where prevailing wage rates are higher.[15] Similarly, some countries give big cities privileged access to specific types of infrastructure grants –particularly those for urban transport— because they recognize that these systems serve the whole nation.  And finally, in education, many countries require local governments that ‘export’ pupils to other jurisdictions to pay them an amount equal to the importing local government’s average per pupil expenditure.

Nonetheless, in most countries where local governments are allowed to either set PIT rates, or in which they derive significant revenue from a PIT share, the general assumption is that big cities can fend for themselves. This is because the revenues these large urban jurisdictions derive from PIT and other taxes are generally so much higher than other local governments –and with them, access to cheap debt capital-- that special measures are deemed unnecessary to support them. Indeed, if special measures are considered necessary, they typically consist of some mix of categorical grants, and taxing back less of their ‘excess’ PIT revenue to fund equalization. 

More important for our purposes, however, is the fact that no European country attempts to provide big cities with additional resources by defining the origin of some –let alone all—of PIT by place employment. Doing so would violate the nature of PIT as a tax as well as the principle that governments should only impose, or in the case of PIT sharing, use the taxes of people who vote for them. And it would run against the basic idea that the fundamental purpose of local governments is to provide services to their residents.

Indeed, of the close to twenty European countries that provide local governments with significant chunks of general revenue through PIT, only one of them –Romania—defines the origin of PIT as people’s place of employment. In Romania, then, the unity of fiscal and political jurisdictions is broken, and some local governments receive the PIT payments of people who do not live in them or vote for their officials. By the same token, some local governments do not get the PIT payments of people who are living in them and who vote for their officials but work elsewhere. Moreover, and as in Ukraine, the allocation of PIT by place of work tends to underfund suburban jurisdictions and over fund urban centers, requiring the equalization system to be larger than necessary.

Nonetheless, there is a crucial difference between Romania and Ukraine: In Romania, multiunit businesses are required to indicate to the tax authorities the actual jurisdictions in which their employees work. As a result, Romania does not have the problem that Ukraine has with large multi-unit businesses sending the tax payments or all their employees to the local governments in which they are legally registered. This is a significant and important difference. Nonetheless, the Romanian system remains an outlier and should not be the model on which Ukraine builds its intergovernmental finance system. Instead, the basic answer to how Ukraine should solve its PIT problem is to do what virtually all European countries that either share PIT with local governments, or give them rate setting powers do, and that is define the legal origin of PIT as a taxpayer’s place of residence.

4. What Might be Done to put PIT on an Origin Basis in Ukraine

As in the rest of Europe, Ukraine should share PIT with local governments on basis of where taxpayers reside. This is the only way to align political and fiscal jurisdictions and to recognize in practice that the fundamental purpose of local governments is to serve the people who vote for them. It is also the only way to eventually give local governments some rate setting powers, and thus further strengthen the linkage between taxation and political accountability. In short, the answer to the question of how Ukraine should share PIT with local governments is straightforward.

The more difficult question is what Ukrainian policy makers need to do to achieve this objective. Here, it is worth beginning with the minimum condition that must be fulfilled to change the existing system: The national government must replace firms as the agent which allocates to local governments their PIT shares. This question is independent of whether these shares are allocated by place of work or place of residence. Instead, it is dictated by the structural requirements of a decentralized order in which the state has agreed to finance local governments by giving them a share of the national taxes generated on their territories. And as the Ukrainian courts have already ruled, this is not a public function that can reasonably be imposed on private firms. This does not mean that private firms should not be responsible for providing important tax information to the national government. They should. They just should not be responsible for determining which local governments gets whose taxes.  

To replace firms as the payment agents of state, the national government must develop the IT and data systems necessary to gather, process, and store information about each taxpayer’s identification number, his or her PIT payments, and the jurisdiction(s) in which he or she may work and/or live so that the State Treasury can send the appropriate PIT shares to the relevant oblasts and hromada budgets. Developing such IT and data processing systems is never easy, not least because they require both inter-ministerial coordination and the restructuring of the internal procedures of key state institutions like the Tax Administration, the Treasury, and the Ministry of Finance. Indeed, it is for these reasons that they rarely succeed if they are not clearly mandated by the highest levels of government and adequately resourced. That said, developing and implementing such systems is not rocket science, and we see no reason why the Ukrainian state should not be able to build such as system by the beginning of fiscal year 2023.

Equally importantly, if one stipulates that the existing system makes no sense, is unfair, inefficient and must be changed, then the costs of moving to a system that allocates PIT by place of residence are roughly the same as moving to one based on place of work. Or at least, they are roughly the same if we do not confuse where people actually live with where they are legally registered to reside, as we will discuss more in a moment. Here, the essential point is that if the costs of moving to either system are similar, then there is no question that policy makers should choose to build a system that allocates PIT by where taxpayers live: This is the only way to unify political and fiscal jurisdictions, and to reap the gains in democratic engagement and accountability that come with it. It is also the only way to eventually decentralize some rate setting powers to local governments.

To fully explain why the costs of moving to either a residency or work-based system of PIT allocation are roughly the same, it is necessary to first do a little recapping, and then to discuss the objections that are typically raised against allocating PIT by place residence in Ukraine. As we shall see, most of these objections fall away if PIT can at least initially be allocated to local governments according to where employees say they are living and not where they are legally registered to reside. Indeed, the very process of getting taxpayers to first reaffirm where they are currently living, and then –if necessary-- to eventually change their legally registered addresses, can be used to kickstart the linkages that origin-based PIT sharing is designed to foster: The accountability linkages between voter/taxpayers and their elected officials; and the financial linkages between local economic development and local budget revenues.

So here is the recap. At present, the Ukrainian state cannot link peoples PIT payments to their actual place of work. Because it cannot do this, it is relying on firms to send people’s PIT payments to the local governments in which they work. And because this is something that firms which employ workers in different places are understandably reluctant to do –and should not be doing— the PIT payments of employees who work outside of a local government in which the firm is legally registered are often being returned to the local government in which the firm is registered –most frequently to Kyiv.

As a result, the Ukrainian state does not know how much PIT revenue Kyiv is getting from workers who neither live nor work in the capital.  Nor does it know how much the local governments in which these employees actually work –and probably live—are losing. What it does know, or at least should know, is that the amounts involved are significant enough to give Kyiv three times as much times PIT revenue per capita as the average local government. Moreover, it does this despite receiving a PIT share one third smaller than all other local governments (40% vs 60%).

Indeed, the fact that Kyiv’s PIT share is one third smaller under the current regime has surprising implications when one looks forward to what it would mean to really allocate PIT by place of work, or for that matter residency. In short, the money that Kyiv is currently getting from taxpayers who neither live nor work in the capital would flow outward to the local governments in which these people actually work and probably live. So, Kyiv’s budget would be hit very hard. In fact, so hard that its PIT share would almost certainly have to be increased to make up for the lost revenue. Moreover, unless the shares of other local governments were lowered, the costs of this upward adjustment would have to be borne by the state budget. This is because the state budget has also been profiting from the fact that it retains 20% more PIT from taxpayers employed --or said to be employed-- in the capital, than it does for taxpayers working elsewhere.

As such, it is fair to say that if the minimal condition for meaningful reform is the state’s investment into the data and IT systems discussed above, the first axiom that follows from this condition is that moving to an origin-based system for allocating PIT is going to require a substantial rethink of both Kyiv’s PIT share and the national budget. And this axiom is true whether we allocate PIT by place of work or place of residence. Or put another way, it is the reallocation of the PIT that is currently being paid to Kyiv for people who neither live nor work in the capital that is going to produce the biggest change in local government budgets, and that this is true whether or not PIT is allocated by place of work or place of residence.

Indeed, if the biggest technical challenge of the next two years is for the state to estimate the financial consequences of the movement to origin-based PIT sharing and then to construct the IT systems necessary for the Treasury to divvy up shares, then the biggest political challenge of the next 24 months is to have an informed discussion over what Kyiv’s PIT share should be; whether the capital should begin to contribute to the equalization system; and how the national government should respond to the loss of revenue that will come –all things being equal—from an increase in Kyiv’s PIT share.

The prospects of such discussions are no doubt troubling for both state officials and Kyiv’s elected leaders. Moreover, they must include the representatives of all hromada because their results will affect the budgets of all local governments. But like it or not these discussions must be held if the current system is to be meaningfully reformed. And like the construction of new IT systems, these discussions must be held whether the goal of the exercise is to allocate PIT by place of work, place of residence, or both.  Or put another way, the only question policy makers really should be asking themselves now is whether the costs of building these IT systems and engaging in these politically unavoidable discussions go up so significantly by allocating PIT by place of residence that they outweigh the benefits in accountability and engagement that will come from aligning political and fiscal jurisdictions.

The answer to this question is no, at least if people’s declared home jurisdictions -and not their legally registered addresses- can be used to allocate PIT for a transitional period. To understand why, it is useful to briefly review the main arguments against allocating PIT by place of residence.  Here, the first thing that needs to be noted is that most of these objections have grown up around the assumption that the current system is based on place of work –which is simply not true. Nonetheless, and until very recently, the Ministry of Finance, the Tax Administration, and Kyiv have all argued against moving to a residency-based system on the grounds that the existing system is based on place of work and functions reasonably well when neither is the case. In short, many of the arguments used against allocating PIT by place of residence have nothing to with the relative virtues of place of residence vs place of work. Instead, they have everything to do with the recognition that moving to either requires a rethink of the existing system, a rethink that neither Ukrainian’s financial ministries nor the representatives of big cities –particularly Kyiv—are thrilled about having.  

The most frequently invoked argument against allocating PIT by residence is that big cities often incur significant costs for commuters. This is true. But it is not really an argument for allocating PIT entirely by place of work. Rather, if anything, it is an argument about the desirability of splitting PIT payments between place of residency and place of work.  And splitting comes with a number of problems.  First, and as already discussed, any allocation by place of work weakens the link between voting and taxpayment and with it the incentives for citizens to hold their local governments accountable. Second, splitting makes it impossible to ever give local governments any rate setting powers because then the officials of one local government could raise the taxes of people who did not vote from them. Third, splitting increases the administrative costs of the system because now data must be maintained on where people both live and work. Worse, people change work more often than they change where they live. So, the difference in the long-term costs of a split system and one based solely on residence are significant. Indeed, it is for all these reasons that no country that we know of allocates PIT by both place of work and place of residence.

Another argument that is making the rounds is that allocating PIT by place of residence incentivizes housing construction --not jobs-- and Ukraine needs jobs more than housing. Sometimes this argument is invoked in support of splitting PIT, and sometimes in support of allocating it entirely by place of work. As an argument for splitting, it fails for the same reasons as discussed above. But it is even more problematic as argument for a system based entirely on place of work. Yes, the origin-based sharing of PIT creates incentives for local government officials to invest in job creation. Indeed, this one of the great virtues of allocating PIT by place of residence. But allocating PIT solely by place of work in the name incentivizing jobs over housing makes no sense: It will starve bedroom communities of the revenues they need to provide basic public services; intensify the need for equalization grants; and breaks all the rules about aligning political and fiscal jurisdictions. In short, there are no good systemic reasons to support either splitting PIT or allocating all of it by place of work.

But there is one reason against allocating PIT by place of residence that must be taken more seriously. The argument is simply that in Ukraine PIT cannot be allocated by place of residence because so many people live in local governments different from the ones in which they are legally registered to reside. And this is true. Indeed, the Ministry of Digitalization estimates that somewhere between 20% and 30% of all Ukrainians live in local governments other than the ones in which they are legally registered. Worse, the argument runs, this cannot be fixed any time soon because there are many obstacles to changing one’s legal place of residence.

Some of these obstacles relate to the fact that changing one’s legal address requires changing all sorts of other documents, a process which can be time consuming and frustrating. Others relate to the way the Ukrainian housing market works: To the reluctance of owners to enter into legal contracts with their tenants out of fear that they will they have difficulty removing them if they don’t pay rent; and to owners fears that they will have to pay taxes on rental income if contracts are legally registered. In short, people argue that not only will it take a long time for people to legally register where they actually live, but until this is done PIT cannot be allocated by place of residence because it will produce radical changes in the distribution of local government revenues.

Moreover, the fear that moving to a residency-based system will produce radical changes in the current distribution of local government revenues is particularly strong in Kyiv and other big cities because it is they that have the most people who rent apartments, but who remain legally registered in other places. In short, Kyiv and other big cities are afraid that moving to a residency-based system will not only entail the loss of the PIT revenues that they now receive from taxpayers who neither live nor work in them, but the loss of the PIT payments of people who live in them but are legally registered elsewhere.

These are not irrational fears. Indeed, once one stipulates that the existing system must be changed, it is only the issues raised by legal residency --the propiska-- that argue against allocating PIT by where taxpayers live. But these arguments, and the fears behind them, can be addressed if the question of how to allocate PIT by where taxpayers live is divorced --at least initially-- from the question of where taxpayers are legally registered. In short, what we are proposing is that for the purposes of allocating PIT, taxpayers simply be asked to declare in which local government they are currently living.  This declaration need not be tied to providing one’s exact address or to presenting a legal rental contract. Nor does it have to be linked to changing other documents required from citizens by the national government. As such, this simple affirmative declaration would sidestep all the obstacles to changing one’s legal registration created by the Ukrainian housing market and by the requirements that different agencies of the national government attach to legal residency. Instead, the declaration would only be used for allocating PIT and for linking –in the minds of citizens—their PIT payments with local governments they elect. In other words, once a decision is made to share PIT by place of residence, both the national government and local governments should actively campaign to get people to declare where they currently live “so that your PIT payments go to the local government that serves you.”

In conclusion, let us sketch out what all this might mean operationally. For starters, legislation should be passed requiring the national government to build the IT systems that would allow the State Treasury to allocate PIT by place of residence by the fiscal year 2023. This legislation should also require the Ministry of Finance to conduct the simulations necessary to rationally review Kyiv’s PIT share and to phase in the new system over three to five years.

Next, the Ministry of Finance must develop a simple electronic format for firms to fill out that includes each taxpayers’ unique identifier, their PIT payments, and at least for the first filing, the jurisdiction in which he or she both works and lives. Most firms have this information already, but for this first filing, they should be asked to request from their employees an affirmative declaration of where they are currently living. At the same time, both the national government and local governments should inform citizens why this information is important and what it will be used for.

Firms will have to bear the costs of providing this information to the national government, but they will not be asked to send the PIT payments of their employees to tax authorities different from those to whom they are already sending them. Instead, the information will be used by the Ministry of Finance to simulate the consequences of moving to a residency-based system, and to build the IT systems necessary for the State Treasury to send PIT payments to the appropriate local governments.

Once the Ministry of Finance has acquired six months of data on PIT payments that are linked to place of work and place of actual residence it can begin to simulate the changes in the distribution of PIT that will be caused by the new system. These simulations will almost certainly show that the biggest shock to the system will have nothing to do with the difference between place of residence and place of work, but with the difference of either and the current system. More importantly, they will allow the Ministry, local governments, and ultimately parliament to think through what Kyiv’s PIT share should look like, whether the city should participate in the reverse grant system, and how the national budget will absorb the losses that will come from having a single PIT-sharing regime for all hromada.

Equally importantly, the simulations will make it possible for the Ministry of Finance to design phase in provisions for the new system. There are many ways this can be done, but the basic idea is that no local government should be allowed to lose or gain more than a certain percentage of revenue a year until the new system is fully phased-in. On the one hand, these sorts of provisions prevent local governments from suddenly losing revenues necessary to pay for basic services, and on the other hand from getting budgetary windfalls that they cannot reasonably consume. In the current situation in Ukraine, however, they will have an additional benefit: they will give local governments time to impress on their voter/taxpayers the importance of correctly indicating where they live, while also perhaps giving state officials new ways to think about how to deal with the problems of both legal registration and the peculiarities of the Ukrainian housing market.

Finally, it is worth underlining that the primary purpose of putting the PIT sharing system on a residency basis is not to radically change the way resources in the system are currently being allocated. Rather, it is about putting the system on a foundation that encourages local government engagement and accountability, and which allows for the automatic adjustment of local government revenues to changes in where people live. Or put another way, Kyiv should understand that any new system will require the upward adjustment of its PIT share, and that all big cities should be fine if –at least initially—PIT is allocated according to where taxpayers say they live and not where they are legally registered. In any case, there is no reason to believe that the costs of moving from the current system, to the one all of Europe uses will be any higher than the costs of moving to the system that until now most Ukrainians thought they had but do not.