By W. B. Kyijomanyi
IntroductionTaxes federal and state/provincial governments can or should levy in a future Federal Uganda
Who shall levy what taxes and how?
The federal and state/provincial governments shall have shared jurisdiction in the following areas:
Exclusive areas of taxation for states/provinces include:
Exclusive taxes for local governments - cities, municipalities, town councils etc.
How could tax compliance be improved in a Federal Uganda?
Fiscal/inter-State competition
In a country with limited manpower and cash what is the most efficient way of sales tax collection?
How does a search for efficiency affect central-provincial power sharing?
How should one address the status of provinces with a very low sales-tax base, by fixed formula or by ad-hoc transfers?
How does the nation address local resistance to imposition of taxes?
Taxation of natural resources
Why equalization?
Key design options
The distributive formula:
Conditionality
SUMMARY AND RECOMMENDATIONS
In this article (several postings on FedsNet), we tried to develop the best taxation model for a future federal Uganda. Taxation is a very technical area, but we tried to make it simple and colourful. Those who have been on this forum for sometime, may recall the efforts made by Mr. Kibuka and others to resolve the expenditure question- “Who does what” in a future Uganda. Although that question has not been fully resolved, we have made some progress to move to another fundamental question-the Revenue assignment, which taxation tries to address. We are trying to find the suitable and applicable model in a future federal Uganda-a Model of Tax Assignment.
The question we are trying to answer is "Who should tax, where, and what?" in a multilevel government in Uganda. Our collective effort is to find answers to the three aspects of this question. The little consensus there is in public economics is the following:
(a) Lower levels of government should as much as possible, rely on benefit taxation of mobile units, including households and mobile factors of production.
(b) To the extent that non-benefit taxes need to be employed on mobile economic units, perhaps for redistributive purposes, this should be done at higher levels of government. e.g. Federal or State government. Only federal and state/provincial governments should impose a corporate and personal income tax. Whether state rates are based on the federal rate or levied on income is a matter to be determined by respective state/provincial governments. I will however make suggestions on what areas in Uganda may apply what criteria. Some states/provincial governments may also decide not to levy income taxes on wages [not all income] as is the case in Texas, Tennessee, Florida, Washington, Nevada, South Dakota, Wyoming, New Hampshire and Alaska. This too will be revisited in latter postings. The federal and state/provincial governments should also be the only levels to levy a sales tax. As it will be explained later, some state/provincial governments may decide not to levy the sales tax, as is the case with some states in the US. We do however, hope that VAT in most jurisdictions (GST in Canada), will be maintained but split between the federal and state/provincial governments at negotiated rates.
(c) To the extent that local governments (cities, municipalities, and town councils) make use of non-benefit taxation, they should employ them on tax bases that are relatively immobile across local jurisdictions e.g. housing. The best candidates for subnational taxes are levies that are (1) on relatively immobile bases (2) when the base is relatively evenly distributed and (3) when yields are likely to be relatively stable. In practice what these amounts to, is that local governments are supposed to primarily rely on user charges and taxes on real property. I believe that property taxes if properly implemented in Uganda have the potential to make up or even surpass tax revenues from poll tax, which should be abolished. I will return to this issue in later postings, but for now it suffices to say that the people likely to be impacted most by a comprehensive and marked assessed property tax regime may fight this tax. They are the same sources defending a poll tax and how unwise it would be to abolish it. Truth be said, they have a vested interest. What is implicit in the arrangements above is that governmental tax objectives are in effect arrayed in a strict hierarchy, in accordance with the subsidiarity approach to assigning expenditures, with those of the federal government taking precedence over those of lower governments. But this is incorrect in a truly federal country, as it will be demonstrated in the next postings.
Taxes federal and state/provincial governments can or should levy in a future Federal Uganda
We have outlined above, the questions we are trying to address in search for an appropriate model for a future federal Uganda. We continue along the same lines but from another perspective. The last part was based on the normative aspects-what ought to be, but this part is rooted in the public choice approach-voting matters so it is better to match revenue with expenditure responsibilities. Governments decide which taxes to impose in terms of a political rather than an economic calculus. This is why there is no flat tax yet in the USA. Governments are more concerned about being re-elected than how efficient the tax systems are. In Uganda's case, you read competing visions on the poll tax, one for total elimination (on both equity and efficiency grounds), while the incumbent was for lower rates. That was a classic case of competition for the tax base-taxpayers/ voters/consumers because it was perceived to affect their electoral chances. What this implies is that taxes assigned to lower governments-local councils/municipalities/cities, something that is usually determined at the discretion [I prefer clear rules over discretion] of higher-level governments fall into one of three categories.
1) Local taxes will be those that are too small to bother with - e.g. Poll tax in Uganda.
2) They may be those that are difficult or costly for the federal [central] government to administer and potentially troublesome. e.g. the poll tax again, but also the property tax.
3) They may be taxes that slip through the cracks such as local business taxes. Essentially, this is a revenue-maximising model (subject to efficiency and equity constraints) which means that state/provincial taxes should be imposed on mobile factors so that competition between such governments (horizontal competition) will limit the grasp of the Leviathan, for example limiting the number of ministers/presidential assistants, from the current high of 60 plus ministers. In Uganda's case, this would mean competition between states because "what is good for the private goose is good for the public gander", that is competition can be healthy and beneficial between governments as between private economic agents. Not long ago, I cited Jinja's demise and how Busoga leaders seem "impotent" to do anything to reverse the town's decline precisely because major decisions are made in Kampala without concern for Jinja's vitality. Horizontal competition between the state/province of Busoga with others would have kicked in had we had a federal system. Vertical competition (competition of governments at different levels) would have saved Jinja's demise. For example, Jinja town might have aggressively moved to either lower or suspend property taxes paid by major industries if that is what it took to compete with Mbarara or Kampala. Nytil is dead because the decisions to allow TAX FREE imports of second hand bales "endiboota" in Uganda were made at the centre. In a federal system Jinja town and the state/province of Busoga would have had a say in how Nytil responded to such unfair competition from "endiboota". Instead good industrial jobs have been lost and with it Jinja's tax base. More devastating however, is the blow to the cotton sector in Uganda. None of these mattered to the folks who unilaterally decided that "endiboota" should be imported into the country tax-free.
In summary, responding to what should be taxed and by whom, state/provincial governments need control their own revenues so to facilitate effective decentralized control of spending. This control requires that states/provinces affect the volume of own revenues significantly at the margin through their own policy choices, in particular by choosing tax rates. That is if, state/provincial governments are expected to act responsibly, they must be able to increase or decrease their revenues by means that make them publicly responsible for the consequences of their actions. To continue with the Jinja example, the state and local governments of Busoga and Jinja respectively would have used their allocated taxation powers in a federal system to do something about Jinja's "demise" or face the wrath of voters at the polls. Unfortunately, because of the unitary system, Jinja will continue to "die" but the local politicians are not likely to pay any costs given lack of political competition in Uganda.
Who shall levy what taxes and how?By now it should be clear that the powers of taxation in a future federal Uganda shall be shared among various levels of governments - federal, state/provincial and local.
The federal and state/provincial governments shall have shared jurisdiction in the following areas:a) Personal Income Tax (PIT) - usually largest revenue earner
b) Corporate Income Tax (CIT) - usually second revenue contributor
c) Retail Sales Tax (RST) - this is currently VAT in Uganda
d) Tobacco Tax
e) Gasoline Tax
f) Fuel Tax - this and the gasoline tax above are usually "dedicated taxes" towards road and infrastructure repairs.
g) Mining Tax
h) Capital gains tax-the Uganda Income tax act will have to be amended, and the amount of capital gains to be included determined. I would however recommend 50% to encourage entrepreneurial risk taking with some non-taxable rewards. Capital gains arise when shares or housing property appreciate in value. 50% inclusion rate would exempt half of such gains from taxation altogether, instead of the current case where f 100% exclusion hence Tumusiime former MD of URC, will or paid no taxes upon selling his cheaply acquired mansion.
i) Exercise and custom dues - while in most federal jurisdictions, such taxes are exclusively allocated to the federal level, this should not be the case in a future federal Uganda. Boarder points (Malaba, Busia, Entebbe, and others) will serve as sources of origin for all taxes levied at such entry points. Accordingly, the current system where such boarder points receive nothing from collected revenues will be eliminated so that a percentage say 2-5% of all customs and exercise levies go to the respective state/province governments where such entry points are located. This will be controversial but most state/provinces have some entry points to benefit. Under this model, the federal government shall not have any exclusive taxing powers.
Exclusive areas of taxation for states/provinces include:1) Land transfer tax-this may be controversial in Uganda where powerful political interests are now some of the largest landowners. This is not a land tax, but a tax on land transfers/transactions, when title deeds change ownership
2) Employer health tax-this is the tax paid by employers towards public health dues. I am assuming that universality and accessibility of health care will be core principles of a future health care act in Uganda. All Ugandans shall be entitled to comparable health care services irrespective of which state/provinces they reside in. Revenues from this tax may be minimal hence the need for federal transfers to all provinces.
3) Any other duties/taxes as may be decided by state/provincial legislatures. This may include among others; succession duties, motor registration fees, license fees, payroll taxes in lieu of personal income tax among certain workers (in many cases, PIT is not paid by taxpayers below a certain income level).
Exclusive taxes for local governments - cities, municipalities, town councils etc.(1) Property taxes - assessed accordingly but preferably market or economic value
(2) Other business taxes as may be decided by elected councils
What this shows is the conventional model of tax assignment, which in effect assigns all significant revenue sources to central/federal governments, would be clearly inappropriate in a future federal Uganda. State/provincial governments will be expected to account for significant proportion of public sector spending, unlike the present decentralization arrangement where districts are simply acting as administrative arms of the centre. If state/provincial governments are to be big spenders and they will have to, they must in the interest of fiscal responsibility and accountability, also be taxes than the current unitarist models currently allows.
How shared taxes shall be collected and divided among Federal and State governments in a future federal Uganda.In all federal jurisdictions the issues outlined in this section are always controversial and are the source of sharp and endless quarrels between governments. Whether taxes will be harmonized and therefore collected by Uganda Revenue Authority on behalf of both federal and state governments; separate and collected by different tax authorities; what taxes may be collected jointly or separately; how tax rates will be determined; what the tax base (what qualifies as income); and if harmonized, how the distributive formula will be determined are issues to be settled through joint federal-state negotiations and agreements. Still I endeavour to offer some suggestions.
It is conceivable that common sense and prudence will prevail to adapt harmonization on efficiency and administrative grounds. Given the dual sovereignty [taxpayers-voters-consumers belong to more than one level of government] in federal arrangements, it makes sense to have one collection agency, and revenues shared among federal and state governments according to an agreed formula. This is the case in most countries for at least the major taxes-PIT and CIT. If that were to happen then state rates on income would be levied in the form of a state surtax (percentage) on the national income taxes. For example, Eastern State may decide to set its rate at say 70% of the federal rate. This would tie state revenues to federal revenues. This arrangement has both merits and demerits. Among the merits are easy of collection. A principal demerit is surrender of autonomy by state governments to the federal government. Moreover, state finances would be dictated by what the federal minister of Finance does in Kampala, meaning that State governments would not have total control on their revenues [budgets]. In Uganda's case, if the state of West Nile were to fully harmonize its tax collections, its rate setting power and independent economic policy making would be tied and determined by what the federal minister does. For example, if the federal minister raises taxes, state taxes would rise accordingly, bringing more revenues to the state treasury; if on the other hand federal taxes are cut, state taxes will proportionally fall, and with them lower revenues. As you can see this would mess up state spending priorities and programs. It is worthy noting that in most federal jurisdictions, poorer or less endowed states usually favour harmonization with the federal government. This is what is referred to as tax on tax.
The other interesting and likely possibility is that given different state needs, and therefore priorities in a future federal Uganda, there won't be consensus among state governments, and between them and the federal government. Accordingly, different states may set different tax rates to raise revenues to fund their respective priorities. States would take their power of tax assignment very seriously to determine the tax rate and base. One of the more contentious issues is defining what income (base) is so various states in Uganda, as it is elsewhere may define it differently. If this were to happen, some states may decide to levy taxes on base. The key merit of this arrangement is that states would be autonomous to set their own tax rates and have control over their state revenues [budgets]. For example if the state of Ankole or Kigezi were to apply this, it would not care very much [though not indifferent] about what the federal finance minister did with taxes in Kampala. If federal taxes were to rise or fall, state taxes would not be affected. This is the arrangement in many of the rich states and provinces in the US and Canada. The point to note here is that such taxes are designed or made within the state/province in the true spirit of federalism. This method may also best serve the interests of states in Uganda with strong cultural institutions such as Buganda, Busoga, Toro, and Bunyoro. Taxes made within state accord them the flexibility to promote and reserve their cultural institutions. The same will apply to areas such as Teso, Nebbi, West Budama with recognized cultural leaders. The other day, there was a story in the papers, how Teso's cultural leader the Emorimor, could not tour areas within Teso because his car had no fuel. The independent rate setting power under this arrangement will allow states to decide which taxes and what rates to levy, something that should help to cater for the welfare and maintenance of cultural leaders. I realize how some are dismissive of international examples, but the French speaking Province of Quebec adapted this arrangement so it can have the critical independence to set its own tax rates and fund its unique programs. This is arrangement is what I called tax on income in my introductory posting. There are however a variety of interesting situations. Uganda as elsewhere may decide to have separate federal and state taxes administered at the state level as in the previous case; joint [harmonized] federal and state taxes administered federally as in the first case in a future Uganda. Although that question has not been fully resolved, we have made some progress to move to another fundamental question-the Revenue assignment, which taxation tries to address. We are trying to find the suitable and applicable model in a future federal Uganda-a Model of Tax Assignment.
The question we are trying to answer is "Who should tax, where, and what?" in a multilevel government in Uganda. Our collective effort is to find answers to the three aspects of this question. The little consensus there is in public economics is the following: (a) Lower levels of government should each state determined by a central formula. These revenues are not really sub national [state] taxes, and may create some tough bargaining between future states and the federal government in Uganda. A case could also be made for maintaining two distinct sales tax bases in which both levels of government tax sales. This may be untidy and probably costly, but it may be argued that such costs should perhaps be viewed simply as part of the price paid for a federal system that presumably has offsetting virtues, such as respecting local state preferences. This is what the Province of Quebec opted to do and may suit Buganda’s interests within Uganda.
Another viable alternative is for both levels of government to maintain independent VATs, perhaps reducing costs through harmonizing bases and to some extent rates-as is now happening in the European Union, while developing a more adequate system to deal with interstate trade. In many federal countries, it is sometimes easier to trade with foreign governments than it is among states. One hopes that in a future federal Uganda, such trade barriers shall not arise. In most US states, it is easier to find Mexican or Canadian beer than beer brewed within some US states. The same applies to Canada, where in some provinces you are more likely to find US or Mexican brewed beer than beer from another province, thanks to NAFTA. Part of this problem arises because states cannot agree whether VAT should be levied on an origin or destination basis. For example, suppose a businesswoman from Mbarara in the state of Ankole or Kigezi were to import cement from Uganda Cement Industry at Tororo in Eastern State, where should the VAT be levied and paid? If it were on an origin basis, VAT would be paid to Eastern State, but if it were on a destination basis, then it would be levied and collected by the State of Ankole or Kigezi. This is so contentious that States and Provinces in the US and Canada can't seem to agree on the method to be employed, hence the interstate/provincial trade barriers within. Producing States would clearly lose revenue if the State sales tax were shifted from an origin to a destination basis. A credible revenue guarantee (from the federal and other States) would be required to make such an approach acceptable.
A final alternative is to let VAT become a joint federal-state tax. Such a tax could be administered by either level of government on a jointly determined base, but with each government determining its own tax rate. From the point of view of fiscal accountability which federalism is really about, this approach is preferable to the Centralized Germany approach. This is how GST is administered in Canada outside the provinces of Quebec and Alberta. This may as well be the model most appealing to States in a future federal Uganda.
Sales tax is such an important, yet controversial tax. To comprehend its complexity, I use Canada, which is perhaps the most interesting country in the world for sales tax aficionados. Canada has several distinct sales tax systems. There is the federal VAT, the good and services tax (GST) that applies throughout the country. As mentioned above, in one province Alberta, the GST is the only sales tax. In four provinces (Ontario, Manitoba, British Columbia and Saskatchewan), in addition to GST, there is a separate RST (retail sales tax of 8%) applied to the GST-exclusive tax base (7%).In the small province of Prince Edward Island, the provincial RST is applied to the GST-inclusive tax base. In three other small provinces (Newfoundland, Nova Scotia, and New Brunswick), there is a joint federal-provincial VAT, called the harmonized sales tax (HST) and administered by the federal government at a uniform rate of 15 percent. In the province of Quebec, there is a provincial VAT, the Quebec Sales Tax (QST), applied to the GST-inclusive tax base. The QST is administered by the provincial government, which also administers the GST in the province on behalf of the federal government.
I cite this example, not because we federalists are taken up with foreign models, but rather to show members that Canada offers a variety of interesting situations: separate federal and provincial VATs administered provincially; joint federal and provincial VATs administered federally; and separate VAT and provincial RST administered separately. A future federal Uganda could learn a lot from the Canadian experience, if we can overcome our "learned helplessness" of that is impossible in Uganda. As Mr. Senyonjo pointed out, with a good tax administration, it is perfectly feasible to operate a VAT at the State level. In principle, it is immaterial whether there are two separate administrations or one, or if there is one, which level operates it. What is critical is either a unified audit or very high level of information sharing. Most important from the perspective of good tax assignment, each taxing government should be able to determine its own VAT rate independently.
There is no doubt that diversity among state policies is a good and efficient virtue. But as Dr. Kigongo noted, states/provinces may engage in un productive competition, by attempting to extract revenues from sources for which they are not accountable, thus obviating the basic argument for their existence. To avoid this danger, it may be desirable to limit state and local governments access to taxes that fall mainly on non residents, such as most natural resources levies [hence the debate as who should have primary jurisdiction-state or central government, as will be flushed out in a later section]; pre retail stage sales tax [hence the controversy as to whether to employ origin or destination criteria, although on efficiency grounds, destination is preferable]; and to some extent non residential property taxes. One way to deal with this problem and minimize un productive competition may be to establish a uniform set of tax bases for local governments, perhaps different rates for different categories such as big cities like Kampala, major towns like Jinja, Masaka, Mbarara, Mbale, Gulu, Lira etc, smaller towns like Tororo, Iganga, Kabale, Mukono, Arua, Soroti etc. and rural areas, with a limited amount of rate flexibility being permitted in order to provide for local effort while restraining un productive competition through the establishment of a floor on rates.
The other controversial case is where you have an "exportable base" as is normally the case with natural resources-gas, oil, is to have a rate ceiling to avoid un warranted exploitation. It is this exploitation that makes ownership of natural resources quite controversial. Those of you who live in the US and Canada may have experienced this during the winter months. States/provinces with substantial oil and gas reserves took in substantial revenues at the expense of consumers in non-gas-oil regions as prices skyrocketed. This is precisely because resources are an "exportable base". Politicians, consumers/taxpayers in Texas, Louisiana or Alberta had little sympathy for the woes facing their counterparts say in the American North East, or Eastern Canada. The higher the prices for gas and oil, the better for the producing states. In fact, in some of the producing states, consumers were rewarded with huge rebates towards their energy bills, thanks to high prices paid by consumers elsewhere. The essence here is that if inappropriate tax bases are assigned to State/provincial governments, wasteful competition and un desirable tax exporting are likely to result. Little wonder then that Texas and Alberta among others have no sales tax. But the problem is not easy to eliminate given the un even geographical distribution of natural resources and the resulting severance of the link between "local taxes" and benefits when the producing states are able to tax such resources beyond any level conceivably justified on environmental or benefit grounds as is often the case. In Uganda’s case, both problems could arise between Jinja and Njeru, located within different states of Busoga and Buganda, so it will have to be carefully considered. But Busoga and Buganda may need to follow the New YORK-New Jersey-Connecticut model where there is co-operation in matters of taxation, since a significant number of people who work in New York City do not live there but in New Jersey and Connecticut. Where should they pay which taxes?
This may arise in Uganda too given the concentrated distribution of oil and gas resources [prospects] around Bunyoro/West Nile region, and has to potential to locate such resources within two different states. Question: if Bunyoro and West Nile can fully exploit their "exportable base", why can't regions with resources like border points do the same? I think they should do so through some agreed formula where a percentage of revenues collected at those points shall be deemed for tax purposes to have been derived from those points (Busia, Malaba, Entebbe and others). The ideal solution is of course to prevent States/provinces with natural resources from doing so, but this would not be politically palatable, as residents from those regions will scream robbery, so it won't be possible. The central government along with the non-resource States/Provinces may have to set up other aspects of intergovernmental finance, particularly transfer systems or one time rebates to off set the resulting distortion as much as possible. More of this to be discussed under taxation of natural resources in a later section, but members should now begin to comprehend the political turmoil in Nigeria and elsewhere.
How could tax compliance be improved in a Federal Uganda?We need to perhaps take a look at Germany's tax laws, which are the most draconian. In Germany, should the revenue agency allege that you owe taxes, you are deemed to owe taxes and must therefore pay the assessed amount in full shortly thereafter, and then litigate to prove your innocence later. It is tough and controversial but Ugandans need some tough medicine in the tax arena, and such a law could move the country towards its taxable capacity [too many people evade taxes because of lax enforcement and lenient penalties]. Those who cared to read on of the links provided by Mr. Senyonjo should know that in the US and Canada among others, tax compliance is voluntary. People don't have to file tax returns with IRS or CCRA unless they owe money to the revenue agencies. Most people however, file returns because they either expect a refund or, need to in order to qualify for all income-based entitlements. The civil penalties for involuntary compliance are usually high-twice the amount owed or some jail term. This cannot work in Uganda, hence the need for some tough tax laws to make tax filing mandatory.
For a start we need the equivalent of a social insurance number; insistence of issuing cash receipts [mandatory], which may require better cash registers, but in the meantime, mandatory receipt books may do. It may also require a mandatory rule whereby transactions beyond a certain amount, must be settled by cheque not cash. Question: in Uganda where politicians and business people are some of the key tax evaders, can a variant of the Germany tax law be passed through parliament? Germany is actually unique in other aspects as well. Even their Chancellor has to pay taxes on any taxable benefits he enjoys. For example, the chancellor must pay taxes on his limos and official residence. Given the extravagance we see in Uganda, several Mercedes, house helpers, drivers, free accommodation etc, is it not time to check such wasteful consumption by re-classifying most of these as taxable benefits? The poor people are harassed to pay poll tax for services that are non-existent, yet our politicians and business leaders are allowed to get away with so may non-taxable benefits.
Fiscal/inter-State competitionContrary to the fears of others, federalism has the potential to un lock Uganda's true economic and political potential. States do not compete through taxes alone, but other channels as well, such as fair business incorporation laws, generous accounting rules etc. Here in the US, the little state of Delaware has curved out a niche in international commerce as the preferred state of incorporation, and is home [headquarter] to some of the biggest and most profitable business conglomerates in the World. Their physical hqs [offices] may be in New York, but they pay their corporate taxes in Delaware not New York. There are states that have created a niche in specific businesses, Connecticut in insurance related businesses, Massachusetts, California and New York for venture capital, Texas for oil etc. All I am trying to show is that states should be given some autonomy through taxes, business incorporation laws or accounting rules to compete for businesses. This is precisely why there is no agreement on what income is because different states/provinces have put into places different accounting rules governing the treatment of issues such as depreciation, capital gains inclusion, deductible business expenses, tax breaks etc. This diversity is good not only for business but also for the public purse. Some may see this as counterproductive but here in the US what qualifies as legal for auditing purposes may not be the case for tax purposes [Not quite two different books but firms try as much as possible to minimize their tax expenses through generous depreciation methods e.g. double declining balance, which is legal for audit purpose, but illegal in some states for tax purpose, because depreciation expense does not involve actual cash flows] some states encourage this more generous method, while others recognize the normal uniform yearly depreciation.
In our case, the state of Karamoja, may decide to become Uganda's Delaware by crafting favourable laws of incorporation, Gulu may decide to use their taxes/business/accounting laws/ to attract firms that have the potential to exploit their expansive land holdings such as cattle ranching; Buganda may decide to use their powers to vie for venture capital or finance related firms. The point I want to end with is federalism offers Uganda a real, not fake chance for creative policy making. As of now, Karamoja cannot attract any substantial business because of some stringent laws requiring corporations to have so many Ugandans on their boards, and some residence requirements, so firms have no incentives to locate outside Kampala/Jinja/Tororo/Mbarara. Such uniform laws are hindering certain and this won't change soon unless we equip such regions with the powers and tools to compete for business and the attendant tax bases that come with. Yet, as we have laboured here, we find it paradoxical that those who make so much noise as to how such regions are ignored are the ones dead set against federalism. Talk of a conspiracy.
In a country with limited manpower and cash what is the most Efficient way of sales tax collection?In most federal jurisdictions, the central government has a comparative advantage in tax collection, so many state/provincial governments contract for its [central government] services in that respect. On efficiency grounds this seems to be the best way forward as long as the state/provincial government (a) decides whether or not to impose a particular tax [in this case VAT], (b) determines the tax rate base, (c) sets the tax rate and, (d) receives all the revenues collected by the central government on its behalf. This is efficient because the only role the central governments plays is that of a collection agent, as done in Canada except Quebec, and is arguably the best system so far in all federal jurisdictions that levy a national sales tax. Why? Because provincial taxing powers by the respective provinces are respected, as they are politically responsible for setting the tax rate. This contrast with the centralized model in Germany where states have no say in how the tax rates are determined, so they don’t quite have taxing powers with regards to VAT. In Uganda's case, the Canadian model would make better sense on both efficiency and accountability grounds, with some different states/provinces adopting whatever arrangement best suits their needs. Some may go for what the big four Canadian provinces did-retain rate setting power [tax powers], while contracting the central government to collect the revenue [on efficiency grounds], some may opt for the Quebec way, set rates and collect revenues for both state and federal level; while others may do what the poor Atlantic provinces did; surrender their rate setting power and collection to the federal government. This is creative policy making at best, something that can only be achieved under federalism. Uganda's needs are diverse, therefore states/provinces should be given the tools and flexibility to make the necessary trade-offs as their legislatures may decide, rather than the one size fits all policy.
How does a search for efficiency affect central-provincial power sharing?Again, the Canadian example best illuminates this question. In any federation, there is always tension between the central and state/provincial governments usually, but not exclusively on powers, including taxation. Germany's centralized model may be more efficient on cost and compliance grounds, but intrudes way too much on the ability of state/provincial governments to chart their own economic policies. Whoever sets the tax rates dictates economic policy. It is not easy and will not be easy in Uganda because many states/provinces will want to exercise their tax assignment power over their own rates [retaining their autonomy], but contracting the central government to collect revenues on their behalf [on efficiency grounds]. This Canadian example shows that a good tax administration system [which will take more than hard work to create in Uganda] can do both: states retain their rate setting power but contract the central government to collect on efficiency grounds. Even the difficult province of Quebec is happy because national-power sharing arrangements are respected.
How should one address the status of provinces with a very low sales-tax base, by fixed formula or by ad-hoc transfers?
Again, the poorer and smaller provinces of Canada (New Brunswick, Newfoundland and Nova Scotia) have demonstrated this. These provinces adopted a joint federal-provincial VAT, called the harmonized sales tax (HST) and administered by the federal government at a uniform rate of 15 percent. Here, the provinces gave up their rate setting power and autonomy to the federal government. Members should also know that their decision was practical in the sense that being smaller, poor and therefore characterised as "have-not", their "viability" depended very much not on local tax bases [not adequate but will soon change as Nova Scotia and Newfoundland have discovered substantial reserves of off shore oil drilling in their waters] for federal transfers through equalization. These three provinces along with Prince Edward Island [smallest province] and Quebec are the net recipients of federal equalization payments/transfers under Canada's "five province standard" [more later under link between taxation and equalization]. In this case, the answer to the question is a combination of both-fixed [uniform rate of 15%] plus transfers through equalization payments which will definitely likely apply in Uganda's case as well, and will have to determined through state-federal negotiations.
How does the nation address local resistance to imposition of taxes?This is applicable to Kampala, the district with the highest taxable base in Uganda, but which nevertheless has a very poor revenue stream. Raising such revenue would entail the taxing of many that are now off the rolls. Take a look at Germany's no nonsense-any due taxes must be paid immediately and litigate later [guilty until proven innocent in tax matters]. Anything less may not be tough enough to deter tax evasion. But in the meantime, the government(s) must try to ensure that taxpayers pay for the services they use [through user fees] and receive the services they pay for [garbage collection, better services in hospitals, well maintained roads] to eliminate the excuses of we pay taxes yet we don't know where our money goes. I certainly believe that a variant of the Germany laws [see posting 9] may be exactly what Uganda needs to minimize tax evasion by politicians and their business allies.
Taxation of natural resourcesBefore, discussing the taxation of natural resources, a brief background comparison between different jurisdictions is order, starting with Nigeria; Canada and the US. Prior to independence, most tax revenue is Nigeria came from trade and commerce, which was mostly concentrated in Northern Nigeria-the present day Hausa/Fulani. The colonial government and politicians from that region agitated for and enacted a law as to how tax revenues among states would be shared. Bear in mind that at that time, oil reserves in Eastern Nigeria had not been discovered. The law stipulated that revenues in pre-independence Nigeria were to be allocated according to the derivation principle [where those revenues were derived or generated], at 100%. 100% of all revenues were retained within the state, while other states received zero. Basically, it was eat what you kill. The 100% formula was intended to encourage tax effort among states. Needless to say, politicians from Northern Nigeria were dead set against bailing out people from Eastern [Igbo and other small groups], or Western State [mostly Yoruba].
Just before independence, prospects for oil in Eastern Nigeria brightened. Accordingly, a law was enacted transferring all natural resources from state to federal jurisdiction, which necessitated an amendment to the derivation formula from 100% local retention to 50%.When oil was discovered, the Eastern region [now subdivided into several states] was to retain 50% of all oil revenues locally, while the other 50% would be shared amongst the federal and other states. That was by any standard very generous to Eastern region. As you might have guessed it, Eastern region's new found wealth did not sit well with political leaders from Northern Nigeria who prior to oil had promoted the 100% formula simply because most revenue was generated within their region. As long as they were benefiting, there was nothing wrong with the 100%derivation formula or arrangement. Shortly after, independence, Nigeria's first civilian government, was overthrown by the military, which then proceeded to whittle down the equalization from 50% local retention to 25%.This did not go well with politicians from the then oil producing Eastern Region.
To cut the whole story shot, you all know about Biafra, and the subsequent plunder Nigeria has gone through. As of now, only 3% of oil revenues are retained by the states, 97% goes to national pool where the federal government takes the largest share, while the rest is shared equally among all states. So Nigeria abandoned the derivation principle in favour of equality among states, where all states irrespective of size receive the same share from the revenue pool. The oil producing states receive 3% plus their share from the pool. As you may have guessed it again, equality of states has created other problems most prominently, demand for statehood, so from three viable regions at independence, Nigeria now has over 30 states, thanks to the equality principle. In Uganda's case, this is the equivalent of every county demanding district representation irrespective of whether such units are viable or not. So in an effort to deprive oil producing states revenue, the Nigerian authorities ended up with more states and demands for statehood, because there is money to be shared equally. Out with derivation, in with equality has created unintended consequences in Nigeria. I urge readers to compare this with Uganda's history.
Another interesting case over ownership is Alberta, Canada. Under Canada's federal constitution, natural resources fall under provincial jurisdictions. Provinces collect royalties from gas, oil, and mining. Federal and state governments levy Corporate Income tax on companies involved in these sectors. Clearly, more revenue goes to the provinces that are entitled to collect royalties plus taxes. As you might have guessed it, this arrangement has not always been smooth. During the 1973 and 1979 oil embargoes, the price of gas skyrocketed in most countries. In Canada, it was a painful experience [recall export the base and need for caps]. Canadians from other provinces except Alberta, where most of the oil and gas resources are located, could not believe that they were paying so much for a resource from within Canada. Many wondered: whose oil is it anyway? The rest of Canada's pain was Alberta's gain. The same could be said of Texas and Louisiana. Higher gas prices created a gang up mentality among the other 9 provinces against Alberta, and pressed the then Federal Government under the late Prime Minister Trudeau to find a Canadian solution against higher gas prices. Under Canada’s constitutional amendment rules, Alberta was outnumbered. The Trudeau, government enacted legislation that allowed the federal government to regulate gas and oil prices, contrary to the constitution. Essentially, the federal government, and the other 9 provinces ganged up against Alberta, which was reaping substantial revenues because of the oil embargo. Albertans never forgave Trudeau and the liberal party for that intrusion on provincial powers. Up to this day, the Liberal party's [Canada’s natural governing party] fortunes in Alberta have never recovered from that painful experience. Only two liberal MPs were elected from Alberta during the last election mostly because of the 70's experience. Again, I invite you to reflect on Uganda's history for any parallels.
The issues then for us are:1) Go the Canadian way and let States/provinces where natural resources fall under provincial jurisdiction, [federal governments still collects CIT].
2) Go the pre-independence Nigerian way and encourage derivation [which was closer to the Canadian principle], grant states/provinces 100% revenues.
3) Go for the immediate post independence Nigeria, and share resources between source states/provinces and federal government plus the rest of states on a 50:50 basis
4) Go the pre-coup Nigeria where derivation was whittled down to 25%, while 75% went to the national pool.
5) Go the current Nigerian way of equality of states/provinces with a paltry 3% going to the source states.
The five possibilities above are intended to guide our dialogue on what may as well be one of the most contentious issue during the federal constitutional negotiations in Uganda: Who shall own natural resources or alternatively how shall the revenues generated from natural resources be shared? Before I proceed to make my practical solution, keep in mind the following four questions. Their answers will help you to reflect on my radical suggestions.
(1) Who does what? The question of expenditure assignment between federal [central] and state/provincial governments. This is really the main variable since the other more or less depends on the answer to this one.
(2) Who levies what taxes? The question of revenue assignment, which my taxation discussion has been trying to answer.
(3) How is any imbalance between the revenues and expenditures of sub-national governments that result from the answers to the first two questions being resolved? The question of vertical imbalance
(4) To what extent should fiscal institutions attempt to adjust for the differences in needs and capacities between governmental units at the same level of government? The question of horizontal imbalance or equalization.
These questions should be approached in the specific circumstances of Uganda. What does federalism aim to achieve? What shall be the main policy objectives of federalism? From what has been floated here, the objectives may include the normal public finance trio of efficiency [allocation], equity [distribution], and stabilization but also economic growth as well as such nebulous but politically resonant goals as regional balance, maintaining national integrity and political stability. Please note that there may be conflicts between these objectives as well as differences between local and central perceptions of the weights to be attached to them. Moreover, as federalists we are cognizant that like all public policies, intergovernmental fiscal policies must take into account both political constraints (such as the strengths of different regions and groups in political decisions) and economic constraints (poor or small tax bases, inadequate financial markets etc). But most important of all, all policy changes and federalism seeks radical changes in Uganda, we must start from where were given our painful history.
As federalists we attach great importance and care deeply about the policy objectives outlined above. I understand the pessimism from the defenders of the status quo, who say that while unitarism has so far failed Uganda miserably, amend it, but do not do away with it. This includes those in Uganda who think decentralization as currently implemented is just fine. Others are of course those who think that unitarism is just fine, so concentrate on job creation for Ugandans. As federalists, we say that both these views are wrong, so rather than amend unitarism, end it, and usher in federalism to rectify the damage and massive failures of unitarism. Many members are aware of the endless rebuttals and suggestions some of us have advanced here. Sometimes we get very passionate, as was perhaps my case with Nytil and how it came to collapse. Responding to the four fundamental questions-questions that apply to all federations irrespective of size or level of development, is and will not be easy in a future federal Uganda. I thought that my brief historical experiences in Nigeria and Canada would sparkle more minds and result in more input. I may not have succeeded in conveying what is at stake when we talk about ownership of natural resources.
In my view, given the answers to the four questions and the objectives outlined above, the questions, we have to grapple with are: Should we apply the derivation principle to natural resources in a future federal Uganda? If the answer is yes, what should the derivation formula be?100%; 50%; 25% or 0%.If the answer is no to derivation then what? Should it be on equality basis-all revenues generated from such resources shall be put into a national pool, shared between the federal government and the states/provinces [equally]. Under this arrangement, natural resources shall be deemed national resources. The third alternative is let such resources fall under state/provincial ownership that shall collect royalties and a portion of their state corporate taxes. The federal government shall not be completely shut out since it too will levy and collect its portion of the corporate income tax. Whether the companies licensed to exploit such resources shall be jointly/privately/publicly owned might not significantly affect generated revenues, but is an issue related to the ownership question, and shall be subject to further negotiation.
Accordingly, on practical and policy considerations, I suggest that ownership of natural resources (oil, gas, gold, copper, cement, diamond etc) shall be governed by the derivation principle. Given that a future central government is expected to maintain services requiring substantial expenditures such as defence, CID, and others, the derivation formula shall take that into consideration. In my view the principle should apply as follows:
25% to home [source] states where the resources are located and revenues derived;
40% to the federal [central government],
35% towards the equalization pool to be shared among the rest of states/provinces not equally but rather on need, population or as may be determined during federal negotiations. My suggestion is attractive because it respects aspects of derivation and equalization, which satisfies the four questions above, the objectives and policy goals of federalism. I have directly addressed the fears of both sides-those who want to see meaningful local revenue retention [federalists] and those who have constantly reminded us about what are we going to do about Karamoja and other disadvantaged states/provinces.
Given the applicability of derivation to natural resources, it is fair to take a similar approach to other revenues sources. In an earlier posting, I suggested that a proportion of customs and other duties collected at Uganda's entry points-boarders, airports, ports should also be deemed to have been derived at those services, and be subject to some form of weak derivation whereby a certain percentage, I urged 2-5% should remain at those sources, but that number could be revised to 5-10% while the rest goes to the federal [central] government. I know this sounds controversial, but why should we choose and pick to apply derivation on natural resources and not other revenue sources? I am for derivation all the way.
In line with this, four principle questions that must be adequately addressed in a future federal Uganda. It should be pointed out that not all federal countries have explicit provisions for equalization [e.g., US}, but practices it through varying levels of transfers, while others such as Canada have such explicit equalization arrangements. Given Uganda's history, we shall to implement equalization transfers for some time, though not forever.
Regardless of the revenue sources [taxing powers] made available to future states/provinces, transfers from the federal [central] to state/provincial governments will undoubtedly constitute an important feature of Uganda's public finances. If, for whatever reason, services [e.g. equivalent national standards in health, education] must be provided by state/provincial government that do not have the fiscal capacity to finance them at levels considered adequate [low or poor tax bases/revenue], if there are externalities associated with the services in question [e.g. pollution from one state to the other], or if a country wishes to take inter-regional differences in needs into account [recall, what shall federalists do to Karamoja, or the lament about everything being within and around Kampala]. This is why transfers through equalization shall be necessary. A well-designed equalization system constitutes an essential component of any decentralization strategy. Note that designing such a system is not easy, much more so given our country's turbulent political history, which must be taken into account when designing the transfer system. A key question is what to focus on effects, or instruments used to achieve them? Bear in mind too that transfers are neither good nor bad. What matters are their effects on such policy outcomes as allocative efficiency [are they being efficiently utilized], distributional equity [are they benefiting the intended groups/services], and macroeconomic stability [are all regions of the country taken care of].
What are critical about intergovernmental transfers is thus not who gives them, who gets them, or what the details of program design are, but solely their effects on policy objectives (emphasis added). What matters are results/outcomes. Properly designed transfers may help rather than hamper sub-national efficiency and accountability even if they finance 90 per cent of local expenditures, while poorly designed transfers will not, even if they finance only 10 per cent of expenditures. Properly designed transfer programs may galvanize local revenue mobilization. It should also be noted that many transfers are intended to achieve broader political goals such as securing and maintaining stability within the federation [rewarding friends and buying off enemies]. These are no means unimportant or irrelevant as some of us who live in federations can testify.
One of Federo's goals is to minimize intervention or orders from the centre. If and when the federal units are agreed on, a committee of experts nominated by both the centre and the units will have to figure out the following: the have (rich) units vs the have-nots (poor). When this is done, an equalisation formula will then be worked out. In essence the have units may not receive any equalisation payments from the centre. The formula must however be flexible to take into consideration future potential (read discovery of oil and other minerals). While Nigerians were discussing the equalisation formula, the Hausa-Fulani regions in the North were opposed to generous equalisation payments to the eastern regions-this is where oil was latter discovered. In the early 60s, the Hausa-Fulani (Northern Nigeria) region generated more revenue than the east. The Hausa-Fulani demanded a more rigid sharing formula that would ensure that most revenue remained where it was derived. As you all know, oil was discovered in Eastern Nigeria and guess what, the Hausa-Fulani wanted the revenue sharing formula changed. The people in the oil-producing region were (still are) not amused. The lesson from Nigeria is Ugandans should be fair in formulating both the equalisation and revenue sharing formulae. I have no idea how wealth will be determined. For example, can Karamoja will all its cattle be considered a have not region? These are the issues to be sorted out. The fact that we shall have the have and have-not units demands a substantial role from the centre. The centre will always have more money than the units. Should the federal government (centre) allocate conditional or block grants? Under the latter, the units get money from the centre and they may do as they please, while under the former, the money is "tied" - if the units don't deliver/use the money for the intended purpose, the centre can penalize the units. This is tricky because the haves would like to see the equalisation payments properly utilised and not squandered on expensive Pajeros, trips etc. by leaders in the have-not units.The key institution for considering the potential sources of revenue for Local Governments should be parliament, well like the House of Representatives in the US. The president should have no arbitrary powers. In many ways, the Ugandan parliament has not been vigilant. In fact, parliament or the present constitution gave the president a blank cheque to spend taxpayers and donors money as he pleases. I understand the president is entitled to a special fund to finance his campaign promises. I gather the same fund finances those selective and sectarian state house scholarships. I realize that given the current arrangement, the centre has "excess" funds and the leaders do not care how such money gets spent. If we have a strong parliament that takes its oversight duties very seriously, such blatant abuse of taxpayer's money would not be happening in Uganda. Luckily under federal, the centre would not have such unlimited revenues.
Federal units should be free to set their own tax rates. The beauty of federalism is that it not only encourages policy innovation, but competition as well. I cannot imagine any federal arrangement where the units are excluded from levying income tax. Clearly there will have to be both a federal and state/provincial income tax. The latter may be levied independently or as a percentage of the federal rate. Under this arrangement, URA will collect the income tax and forward the state/provincial portion. URA is well qualified since there is one taxpayer.It is also possible that the units will decide on any additional revenues as long as they are within their powers. Units may decide to levy other taxes such as sales and property. I believe once federo is achieved, many units will decide to do away with the much hated graduated/poll tax. The point here is that units should have the flexibility to determine their tax structures depending on their budgetary needs. If the government of Karamoja decided to waive any taxes to lure investment, that is a decision they should make, not the centre.
State/provincial governments should be free to borrow from capital markets anywhere. I see no reason why state/provincial governments should not issue bonds to finance legitimate projects. We should move away from the mind set that the centre knows best. If the elected government of Karamoja decides to issue local bonds to improve their local roads, they should not be constrained by a centre based board-a board whose members may never have been to Karamoja and may therefore not be familiar with local needs. Key design optionsThree key factors in the design of intergovernmental fiscal transfers are the size of the 'distributable pool, the basis for distributing transfers and conditionality [to minimize collateral damage to efficiency objectives of federalism for example]
1) Determining the distributable pool is an important characteristic of any good system of intergovernmental transfers [grants] is stability. Karamoja shall know that such equalization cheques will be delivered for several years. Another is flexibility. How can these apparently contradictory characteristics be achieved simultaneously? Basically, there are only three ways to determine how much money shall be distributed through intergovernmental fiscal transfers
(a) As a fixed proportion of central government revenues or some other basis, e.g. as percentage of GDP
(b) On an ad-hoc basis, that is, in the same way as any other budgetary expenditures
(c) On a 'formula-driven basis for instance, as proportion of specific local expenditures, or in relation to some general characteristic of the recipient jurisdiction [greater need]. I personally would prefer this because the rules are more or less clear.
A better way to provide both some degree of stability to the states/provinces, and some degree of flexibility to the central government is by establishing a fixed percentage of all central taxes (or current revenues) to be transferred. Sharing specific national taxes is less desirable than sharing all national taxes because over time central governments will understandably tend to increase more those taxes which they do not have to share [hence my suggestion of no exclusive taxes to the central government]. Without going into the more complicated and dry details, it may be desirable to base the total amount transferred on a more stable macro economic measure such as a moving average of GDP growth [to avoid the impact of external shocks, like low oil and export prices]. This is the case in Canada where transfers are based on the five province standard [Ontario, Quebec, Manitoba, British Columbia and Saskatchewan], the average of these determines how transfers are distributed. Note that the poorest and richest provinces are not included in formulating the national standard on sound empirical grounds. In Uganda's case, whether we adopt a standard based on a few representative states/provinces, or whether we adopt a national standard based on the average of all states/provincial standards is something that will have to be determined during future federal negotiations in Uganda. Members should however, note that this is a very controversial issue. In Canada's case, of the 5 provinces that constitutes the standard, Ontario and British Columbia do not receive any transfers from the equalization pool; Manitoba and Saskatchewan may or may not receive anything depending on wheat prices, while Quebec is the only one that always receives transfers. Those of you who have been here long can understand why. This may as well be the case in Uganda.
The distributive formula:A sound transfer system distributes funds among recipient jurisdictions on the basis of a formula. Discretionary or negotiated transfers are undesirable. The essential ingredients of most formulas are needs and capacity. Needs may be roughly proxied by some combination of population and other characteristics. A more difficult, but conceptually critical, problem is to include a measure of the capacity of states/provinces to raise resources, given the revenue authority at their disposal [recall concerns about destructive competition where some states may leave money at the table]. In Uganda's case, the best structure might be to provide each state with sufficient funds [own-source plus transfers] to deliver a centrally predetermined level of services [e.g., health care]. Because capacity based transfers are in principle based on measures of potential revenue raising capacity (not actual revenues), no disincentive to fiscal efforts is created by this approach. As you all know, there are differentials in needs [Karamoja, Gulu/Kitgum Vs the rest] and in the cost of providing services (for example in rural or densely populated areas), which should be taken into account in such formulas if desired.
ConditionalityOnce the total amount to be distributed has been decided, and the basic distribution formula determined, the key remaining question is whether the transfer should be made conditional on the provision of certain services at specified levels [e.g., health care, education]. Two quite different approaches may be taken. On one hand, to the extent the primary objective of transfers is to ensure that all regions of the country expected to provide services at acceptable minimum standards [not equal] have adequate resources to do so; simple 'lump-sum ' transfers with no conditionality other than the usual requirements for financial auditing will suffice. The assumption here is that the flow of funds to locally responsible political bodies will ensure sufficient accountability and that it is neither necessary nor desirable for the central government to attempt to interfere with, or influence local expenditures choices. This is the position state/provincial rights defenders prefer. On the other hand, if the central government is in effect using state/provincial governments as agents in executing some national policies such as providing universal primary education at the specified level throughout the country, then it would seem to make sense to make the transfer conditional upon the funds actually being spent on education or on the attainment of some level of educational performance [not purchase of say Pajeros]. Bear in mind that money is fungible and this may not always be possible.
There is expenditure conditionality, which ensures that grant funds are spent on the specified service. But does not guarantee that funds do not simply displace "own” revenues that would otherwise have been spent upon the service in question [buy Mercedes Benzes with local funds in the knowledge that federal money is on the way]. In addition, performance conditionality focuses on outputs rather than inputs-for example, the proportion of students achieving certain standards rather than the amount spent on education [UPE in Uganda]. I should say that this approach has considerable merit if well implemented. It focuses on what is presumably the real policy objective: education.
Equalization grants serve two important and distinct rationales. The first rationale, and the most important is to provide the necessary underpinning for decentralization in general, by equalizing to some level the fiscal capacity of state/provincial departments. The basic economic case for such general-purpose (unconditional) transfer is to enable poorer states/provinces to respond adequately to central transfers intended to generate the appropriate level of public goods. A second more, political rational in Uganda's case is to provide sufficient resources to enable all sub-national governments, even the smallest and poorest to provide a basic package of local services other than health, education, or construction of infrastructure. A federal Uganda will need to establish a much better tax regime given the current state in these areas.
How shall fiscal transparency, which is fundamental to sound public policy, be enforced? Whether Uganda decides to set up an intergovernmental fiscal agency to produce periodic reports on the State of the federation, is something negotiators will have to deal with.
SUMMARY AND RECOMMENDATIONSThis is the summary and recommendations of how taxation - the problem of Tax Assignment in a future federal Uganda should be handled, as put across in postings one to twelve. One question that will have to be determined is whether Uganda opts for the source or resident principle in future tax assessments. It is trade-off negotiators will have to grapple with, given the fact that a sizeable number of productive Ugandans live abroad; some of who would like to see a change in policy towards dual citizenship. The rest is an edited version of what was posted earlier. Postings on taxation tried to link the following four questions that all federations grapple with.
(1) Who does what? The question of expenditure assignment between federal [central] and state/provincial governments. This is really the main variable since the others more or less depend on the answer to this one.
(2) Who levies what taxes? The question of revenue assignment, is what these postings tried to answer.
(3) How is any imbalance between the revenues and expenditures of sub-national governments that result from the answers to the first two questions being resolved? The question of vertical imbalance
(4) To what extent should fiscal institutions attempt to adjust for the differences in needs and capacities between governmental units at the same level of government? The question of horizontal imbalance or equalization.
These questions help us to understand the following: What does federalism aim to achieve? What shall be the main policy objectives of federalism? From what has been floated here, the objectives may include not only the normal public finance trio of efficiency [allocation], equity [distribution], and stabilization but also economic growth as well as such nebulous but politically resonant goals as 'regional balance' and maintaining national integrity and political stability. Moreover, as federalists we are cognizant that like all public policies, intergovernmental fiscal policies must take into account both political constraints (such as the strengths of different regions and groups in political decisions) and economic constraints (poor or small tax bases, inadequate financial markets etc).
The question these postings tried to answer is one famously posed by Professor Richard Musgrave: "Who should tax, where, and what?" in a multilevel government in Uganda. Our collective effort is to find answers to the three aspects of this question. The little consensus there is in public economics is the following:
(a) Lower levels of government should as much as possible, rely on benefit taxation of mobile units, including households and mobile factors of production.
(b) To the extent that non-benefit taxes need to be employed on mobile economic units, perhaps for redistributive purposes, this should be done at higher levels of government. e.g.. Federal or state governments. Whether state rates shall be a surtax on the federal rate or levied on income is a matter to be determined by respective state/provincial governments
(c) To the extent that local governments (cities, municipalities, town councils) make use of non-benefit taxation, they should employ them on tax bases that are relatively immobile across local jurisdictions e.g.. housing. The best candidates for sub national taxes are levies that are (1) on relatively immobile bases (2) when the base is relatively evenly distributed and (3) when yields are likely to be relatively stable. In practice what these amounts to is that local governments are supposed to primarily rely on user charges and taxes on real property. I believe that property taxes if properly implemented in Uganda have the potential to make up or even surpass tax revenues from poll tax, which should be abolished.
In responding to what should be taxed and by whom, state/provincial governments need to control their own revenues in so as to facilitate effective decentralized control of spending. This control requires that states/provinces affect the volume of own revenues significantly at the margin through their own policy choices, in particular by choosing tax rates. That is if, state/provincial governments are expected to act responsibly, they must be able to increase or decrease their revenues by means that make them publicly responsible for the consequences of their actions.
The federal and state/provincial governments shall have shared jurisdiction in the following tax areas:
a) Personal Income Tax (PIT) - usually largest revenue earner
b) Corporate Income Tax (CIT) - usually second revenue contributor
c) Retail Sales Tax (RST) - this is currently VAT in Uganda
d) Tobacco Tax
e) Gasoline Tax
f) Fuel Tax - this and the gasoline tax above are usually "dedicated taxes" towards road and infrastructure repairs.
g) Mining Tax
h) Capital gains tax-the Uganda Income tax act will have to be amended, and the amount of capital gains to be included determined. I would however recommend 50% to encourage entrepreneurial risk taking with some non-taxable rewards.
Exercise and custom dues-while in most federal jurisdictions, such taxes are exclusively allocated to the federal level, this should not be the case in a future federal Uganda. In accordance with the derivation principle, taxes collected at boarder points (Malaba, Busia, Entebbe, and others) shall be deemed to have been derived at such entry points for the purposes of taxation. Accordingly, the current system where such boarder points receive nothing from collected revenues will be eliminated so that a percentage say 2-5% of all customs and exercise levies go to the respective state/province governments where such entry points are located. Under this model, the federal government shall not have any exclusive taxing powers.
Exclusive areas of taxation for states/provinces shall include:
1) Land transfer tax - this may be controversial in Uganda where powerful political interests are now some of the largest landowners. This is not a land tax, but a tax on land transfers/transactions, when title deeds change ownership
2) Employer health tax - this is the tax paid by employers towards public health dues.
3) Any other duties/taxes as may be decided by state/provincial legislatures. This may include among others; succession duties, motor registration fees, licence fees, payroll taxes in lieu of personal income tax among certain workers.
Exclusive taxes for local governments-cities, municipalities, and town councils.
(1) Property taxes - assessed accordingly but preferably market or economic value
(2) Other business taxes as may be decided by elected councils. Local governments may also charge the following user charges (you use the service you pay, you don't use the service you don't pay)
(a) Parking charges
(b) Any other use fees as may be determined by the elected bodies/councils. State/provincial governments shall also be encouraged to levy user charges to finance government expenditures. The following charges have promise:
(1) Road toll
(2) Car tax
Whether taxes will be harmonized and therefore collected by Uganda Revenue Authority on behalf of both federal and state governments; separate and collected by different tax authorities; what taxes may be collected jointly; or separately; how tax rates will be determined; what the tax base; and if harmonized, how the distributive formula will be determined are issues to be settled through joint federal-state negotiations and agreements. The general sales or VAT will perhaps be the most complicated and contentious to levy in a future federal Uganda. Most federal countries except the US have figured that the simplest practical way to run a federal-state sales tax [VAT] system is to adopt a form of revenue sharing based on an agreed formula, given the high administrative and compliance costs. In the US, the sales tax (RST-regional sales tax) is only collected at the state level, making the US the only country without a sales tax at the federal level.
In some federal countries, VAT may be levied only at the central level. Germany for example, has a single VAT levied at the national level, although a proportion of VAT revenue is shared on a formula basis with the states. Among other federal countries, Switzerland also has a VAT only at the central level. Australia too has only a federal VAT. This may also be the best model for Uganda in terms of administrative and compliance costs. Only Canada, Brazil, India, and Argentina currently attempt to tax sales at both state and federal levels, and only Brazil and a few provinces in Canada (Quebec) attempt to levy a VAT at both levels. Canada has several distinct sales tax systems. There is the federal VAT, the good and services tax (GST) that applies throughout the country. As mentioned above, in one province Alberta, the GST is the only sales tax. In four provinces (Ontario, Manitoba, British Columbia and Saskatchewan), in addition to GST, there is a separate RST (retail sales tax of 8%) applied to the GST-exclusive tax base (7%).In the small province of Prince Edward Island, the provincial RST is applied to the GST-inclusive tax base. In three other small provinces (Newfoundland, Nova Scotia, and New Brunswick), there is a joint federal-provincial VAT, called the harmonized sales tax (HST) administered by the federal government at a uniform rate of 15 percent. In the province of Quebec, there is a provincial VAT, the Quebec Sales Tax (QST), applied to the GST-inclusive tax base. The QST is administered by the provincial government, which also administers the GST in the province on behalf of the federal government.
I cited this example, not because we federalists are taken up with foreign models, but rather to show that Canada offers a variety of interesting situations: separate federal and provincial VATs administered provincially; joint federal and provincial VATs administered federally; and separate VAT and provincial RST administered separately. A future federal Uganda could learn a lot from the Canadian experience in implementing a shared VAT between central and state/provincial governments. There is no doubt that diversity among state policies is a good and an efficient virtue. But as noted, states/provinces may engage in unproductive competition, by attempting to extract revenues from sources for which they are not accountable, thus obviating the basic argument for their existence. To avoid this danger, it may be desirable to limit state and local government access to taxes that fall mainly on non-residents, such as most natural resources levies. One way to deal with this problem and minimize un productive competition may be to establish a uniform set of tax bases for local government, perhaps different rates for different categories such as big cities like Kampala, major towns like Jinja, Masaka, Mbarara, Mbale, Gulu, Lira etc, smaller towns like Tororo, Iganga, Kabale, Mukono, Arua, Soroti etc. and rural areas, with a limited amount of rate flexibility being permitted in order to provide for local effort while restraining un productive competition through the establishment of a floor on rates.
There is the case where you have an "exportable base" as is normally the case with natural resources such as gas and oil. The solution is to have a rate ceiling to avoid un warranted exploitation. The essence here is that if inappropriate tax bases are assigned to State/provincial governments, wasteful competition and un desirable tax exporting are likely to result. But the problem is not easy to eliminate given the un even geographical distribution of natural resources and the resulting severance of the link between "local taxes" and benefits when the producing states are able to tax such resources beyond any level conceivably justified on environmental or benefit grounds as is often the case. This may arise in Uganda too given the concentrated distribution of oil and gas resources [prospects] around Bunyoro/West Nile region, and has to potential to locate such resources within two different states
Based on practical and policy considerations, it was suggested that ownership of natural resources (oil, gas, gold, copper, cement diamond etc) shall be governed by the derivation principle [assuming that surface rights will be harmonized with sub surface revenues in favour of land owners/states], whereby revenues shall be shared as follows:25% to home [source] states where the resources are located and revenues derived;
40% to the federal [central government];
35% towards the equalization pool to be shared among the rest of states/provinces not equally but rather on need, population or as may be determined during federal negotiations.
Key design optionsThree key factors in the design of intergovernmental fiscal transfers are the size of the distributable pool, the basis for distributing transfers and conditionality [to minimize collateral damage to efficiency objectives of federalism for example]
1) Determining the distributable pool an important characteristic of any good system of intergovernmental transfers [grants] is stability. Karamoja shall know that such equalization cheques will be delivered for several years. Another is flexibility. How can these apparently contradictory characteristics be achieved simultaneously? Three ways to determine how much money shall be distributed through intergovernmental fiscal transfers are:
(a) As a fixed proportion of central government revenues or some other basis, e.g. as percentage of GDP
(b) On an ad-hoc basis, that is, in the same way as any other budgetary expenditures
(c) On a 'formula-driven basis' for instance, as proportion of specific local expenditures, or in relation to some general characteristic of the recipient jurisdiction [greater need]. This would be preferable since the rules are more or less clear.
A better way to provide both some degree of stability to the states/provinces, and some degree of flexibility to the central government is by establishing a fixed percentage of all central taxes (or current revenues) to be transferred. Sharing specific national taxes is less desirable than sharing all national taxes because over time central government will understandably tend to increase more those taxes which they do not have to share [hence my suggestion of no exclusive taxes to the central government]. It may be desirable to base the total amount transferred on a more stable macro economic measure such as a moving average of GDP growth [to avoid the impact of external shocks, like low oil and export prices]. This is the case in Canada where transfers are based on the five-province standard [Ontario, Quebec, Manitoba, British Columbia and Saskatchewan], the average of these determines how transfers are distributed. Note that the poorest and richest provinces are not included in formulating the national standard on sound empirical grounds. In Uganda's case, whether we shall adopt a standard based on a few representative states/provinces, or whether we adopt a national standard based on the average of all states/provincial standards is something that will have to be determined during future federal negotiations.
The distributive formula:
A sound transfer system distributes funds among recipient jurisdictions on the basis of a formula. Discretionary or negotiated transfers are undesirable. The essential ingredients of most formulas are needs and capacity. Needs may be roughly proxied by some combination of population and other characteristics. A more difficult, but conceptually critical, problem is to include a measure of the capacity of states/provinces to raise resources, given the revenue authority at their disposal. In Uganda's case, the best structure might be to provide each state with sufficient funds [own-source plus transfers] to deliver a centrally predetermined level of services [e.g., health care]. Because capacity based transfers are in principle based on measures of potential revenue raising capacity (not actual revenues), no disincentive to fiscal efforts is created by this approach. There are differences in needs [Karamoja, Gulu/Kitgum Vs the rest ] and in the cost of providing services (for example in rural or densely populated areas), which should be taken into account in such formulas if desired.
Conditionality
Once the total amount to be distributed has been decided, and the basic distribution formula determined, the key remaining question is whether the transfer should be made conditional on the provision of certain services at specified levels [e.g., health care, education]. On the other hand if the central government is in effect using state/provincial governments as agents in executing some national policies such as providing universal primary education at the specified level throughout the country, then it would seem to make sense to make the transfer conditional upon the funds actually being spent on education or on the attainment of some level of educational performance [not purchase of say Pajeros]. In addition, performance conditionality focuses on outputs rather than inputs-for example, the proportion of students achieving certain standards rather than the amount spent on education [UPE in Uganda]. This approach has considerable merit if well implemented. It focuses on what is presumably the real policy objective: education.
As it may be now clear, equalization grants serve two important and distinct rationales. The first rationale, and the most important is to provide the necessary underpinning for decentralization in general, by equalizing to some level the fiscal capacity of state/provincial departments. The basic economic case for such general-purpose (unconditional) transfer is to enable poorer states/provinces to respond adequately to central transfers intended to generate the appropriate level of public goods. A second more, political rationale in Uganda's case is to provide sufficient resources to enable all sub-national government, even the smallest and poorest to provide a basic package of local services other than health, education, or construction of infrastructure. A federal Uganda will need to do a much better job given the current state in these areas. To ensure fiscal transparency, which is fundamental to sound public policy, negotiators may decide whether or not to set up an intergovernmental fiscal agency to produce periodic reports on the State of the federation.
To address local resistance to imposition of taxes in Uganda, it was recommended that Uganda take a hard look at Germany's no nonsense tax laws. Anything less may not be tough enough to deter tax evasion. But in the meantime, the government(s) must try to ensure that taxpayers pay for the services they use [through user fees] and receive the services they pay for [garbage collection, better services in hospitals, well maintained roads] to eliminate the excuses of we pay taxes yet we don't know where our money goes. A variant of the Germany laws may help to inculcate a sense of duty towards tax payment in Uganda, and thus, minimize tax evasion by politicians, and their business allies.
