SLIDE 1
ECON 551: Lecture 10 2 of 40
Agenda
- 1. Overview
- 2. Models of Federalism: Breton + Oates.
- 3. Intergovernmental Grants
- 4. Equalization Grants
Econ 551 Government Finance: Revenues Fall 2019 Given by Kevin - - PowerPoint PPT Presentation
Econ 551 Government Finance: Revenues Fall 2019 Given by Kevin - - PowerPoint PPT Presentation
Econ 551 Government Finance: Revenues Fall 2019 Given by Kevin Milligan Vancouver School of Economics University of British Columbia Lecture 10: Federalism and Intergovernmental Grants ECON 551: Lecture 10 1 of 40 Agenda 1. Overview 2.
SLIDE 2
SLIDE 3
ECON 551: Lecture 10 3 of 40
Overview of Federalism
What should be the relationship between national and subnational jurisdictions? Should there even be subnational jurisdictions? How many? Three big ones? Many little ones? How would a non-economist motivate subnational jurisdictions? Allows political liberty and cultural autonomy. Schools, church, language. Think of Switzerland, Canada, Germany. Fosters political participation
SLIDE 4
ECON 551: Lecture 10 4 of 40
Economic Motivations for Federalism
Local Public Goods: the MES for public goods might be small. Spillovers: The extent and nature of spillovers among regions Laboratory: Justice Brandeis in New State Ice Co. V. Liebman 285 US 262, 311 (1932). “It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.”
SLIDE 5
ECON 551: Lecture 10 5 of 40
In 1926, the shares were: 37.8 fed, 20.2 prov, 42.0 local. (Source: Musgrave, Musgrave, and Bird)
SLIDE 6
ECON 551: Lecture 10 6 of 40
SLIDE 7
ECON 551: Lecture 10 7 of 40
SLIDE 8
ECON 551: Lecture 10 8 of 40
Agenda
- 1. Overview
- 2. Models of Federalism: Breton + Oates.
- 3. Intergovernmental Grants
- 4. Equalization Grants
SLIDE 9
ECON 551: Lecture 10 9 of 40
Breton (1965): “A Theory of Government Grants”
Theory hinges on the provision of public goods. Which level of government should provide them? Breton describes public goods with various degrees of ‘localness’. There are International goods (Clean atmosphere, oceans, defence alliances) National goods (legal system, defence) Provincial goods (some resource management) Local goods (parks) Private goods (my lunch)
SLIDE 10
ECON 551: Lecture 10 10 of 40
Breton (1965): “A Theory of Government Grants”
In a world like this, Breton describes the ‘optimum constitution.’ All objective benefits of the local good are exhausted within the border of the jurisdiction. Provides a ‘perfect mapping’ between the scope of local public goods and the political jurisdiction. This generalizes the Tiebout idea to vertical levels of goods. With lump-sum or benefit taxes, we get Pareto optimal allocations. OR, with a supra level of government, they could make conditional grants to ensure that each level can pay for its optimal level of goods.
SLIDE 11
ECON 551: Lecture 10 11 of 40
Solutions to inter-jurisdictional externality problem
How can we solve this externality problem? Try central government. Could be direct provision by feds. Could be quantity control through regulations or conditional grants. Could be price control through subsidies / matching grants. Could we just let Coasian bargaining solve the inter-jurisdictional problem? Translink presumably accounts for city-to-city spillovers. Garbage collection and fire stations, however, are done locally. If there are spillovers, then they made a deal to account for them in that area alone but not for all goods. But, bargaining problems: uncertain property rights, uncertainty about
- thers’ threat points, free-riding by jurisdictions, enforceability of
agreements—governments might renege.
SLIDE 12
ECON 551: Lecture 10 12 of 40
Oates and ‘fiscal federalism’
The seminal work in this area is a book by Wallace Oates called Fiscal Federalism (1972). The model laid out in the book studies the costs and benefits of decentralization. Local Governments: Can respond to local tastes and preferences. Cannot produce public goods efficiently if MES is large, or spillovers exist. Central Governments: Can deal with externalities and scale. But cannot respond to local tastes—assumption of “policy uniformity.”
SLIDE 13
ECON 551: Lecture 10 13 of 40
The Oates ‘Decentralization Theorem.’
“For a public good—the consumption of which is defined over geographical subsets of the total population, and for which the costs of providing each level of output of the good in each jurisdiction are the same for the central of the respective local government—it will always be more efficient (or at least as efficient) for local governments to provide the Pareto-efficient levels of
- utput for their respective jurisdictions than for the central government to provide any specified and
uniform level of output across all jurisdictions” Here is my restatement of it: (1) if a public good costs the same if provided by either level of government, (2) And the good must be provided uniformly if provided centrally, (3) And jurisdictions are formed according to a Bretonian ‘optimum constitution’ Then, local government provision is weakly more efficient than central government provision.
SLIDE 14
ECON 551: Lecture 10 14 of 40
The Oates ‘Decentralization Theorem.’
The central tension is between policy uniformity and externalities. If no externalities, then local provision is weakly preferred. If all had same tastes, then central provision is weakly preferred. Optimal division is found at the point where the cost of policy uniformity equals the gain from internalizing the externalities. Think of this in a two-by-two box: Same tastes Different tastes Spillovers Centralization Tension No spillovers Either / or Decentralization
SLIDE 15
ECON 551: Lecture 10 15 of 40
Agenda
- 1. Overview
- 2. Models of Federalism: Breton + Oates.
- 3. Intergovernmental Grants
- 4. Equalization Grants
SLIDE 16
ECON 551: Lecture 10 16 of 40
Government grants
A good reference is the Boadway and Wildasin text book listed on the syllabus. Also, Handbook of Public Economics Chapter 11 by Rubinfeld. Imagine that we are living in Breton’s world of an ‘optimum constitution.’ Spending on public goods is contained within jurisdictional borders. Now imagine that the tax revenue to pay for the goods cannot be efficiently raised within each jurisdiction. Perhaps raising revenue on mobile factors more efficiently done nationally. In this case, we would have a vertical fiscal gap between the spending responsibilities and the revenue raising responsibilities of some order of government. This generates the need for grants between levels of governments. What does this look like in Canada?
SLIDE 17
ECON 551: Lecture 10 17 of 40
SLIDE 18
ECON 551: Lecture 10 18 of 40
Who should raise revenue?
Why centrally? Economies of scale: Fixed costs can be spread over more bodies. However, in Canada 9 out
- f 10 provinces set their own rates but contract the CRA to collect it on their behalf.
Fiscal externalities: With mobility, attempts at redistribution will be frustrated. As you move up to higher-tier jurisdictions, mobility becomes less easy, meaning that the efficiency cost of taxation is lower. Why locally? Accountability: easier for citizens to follow the dollars. Preferences: some places may want different patterns of taxation. Efficiency: income distributions may differ across provinces.
SLIDE 19
ECON 551: Lecture 10 19 of 40
Conditional vs. Unconditional, matching vs. block
There are many different ways one could imagine the transfers taking place between levels of government. At one extreme, you could imagine sending a ‘blank cheque’ along to the other level of government. At the other extreme, you could imagine a grant that ‘must’ be spent on one type of government program. In addition, there could be a ‘match’ rate; for every $1 of lower level spending, the higher level government will match with a $1. We will look at different kinds of structures for grants Conditional vs. unconditional block grants. Matching grants.
SLIDE 20
ECON 551: Lecture 10 20 of 40
Unconditional Block Grant
Let’s start with looking at the most simple case: a non-matching block grant. In Canada, this is the relevant case for the Canada Health and Social Transfer (since 1996) and the Equalization program.
After tax income Public Good U0 U1 A B
SLIDE 21
ECON 551: Lecture 10 21 of 40
Unconditional Block Grant
The graph shows: After tax income of the province on the vertical axis. (This could equivalently be thought of as amount of the private good.) Spending on the public good on the horizontal axis. ‘Community Indifference Curves’ representing the ‘province manager’ or the median voter. The effect of the block grant is to move the budget constraint out in a parallel way. The equilibrium moves from A to B. Point B could be to the left or to the right of point A, so that the block grant could lead to higher or lower public good spending. It depends if the public good is normal or inferior. Bradford and Oates (1971) show that grants should almost completely crowd out local government spending, because the median voter takes the grant as an increase of income. The MPC on local public goods is likely only around 5%, so we shouldn’t expect a large increase in spending.
SLIDE 22
ECON 551: Lecture 10 22 of 40
Flypaper effect
How much do grants increase spending? Do they ‘stick’ where they land? What are some explanations for this phenomenon? Econometric explanation: Unobserved variables. High spending states might attract more grants. Fiscal Illusion: The spending decisions depend on the source and perceptions by voters of the
- revenue. Grants are not a ‘veil’ for a federal tax cut. This might occur if voters are
imperfectly informed, and/or bureaucrats try to maximize the size of their bureaus. (Niskanen 1971) Counterexample: late 1990s Canada Federal government increased CHST Provinces (e.g. Ontario) started to cut taxes. Feds whined that provinces weren’t spending on health.
SLIDE 23
ECON 551: Lecture 10 23 of 40
Conditional Block Grant
Imagine that the federal government now decides that they don’t like the spending decisions of the subnational government and wants to influence their decisions. Why might the feds choose a different level of spending than the subnational government? Different political preferences. Not the best reason – why impose the political preference of the federal voters on the voters of one province. Externalities. This is a better reason. If providing a program had externalities, then the feds might be able to better account for them than could any individual provinces. For example, provinces might provide too small welfare programs because they are worried about mobility. Migrants would come to take advantage of their benefits, so the provinces’ equilibrium welfare rates are too low.
SLIDE 24
ECON 551: Lecture 10 24 of 40
What does a conditional block grant look like?
Initial equilibrium at C. Conditional grant means that you essentially are given an extra endowment of the Public Good, moving the budget constraint out horizontally. If the local government just spent the new money on the desired good, then E. HOWEVER, the local government re-
- ptimizes and ends up at D – part of the new
endowment gets consumed as tax cuts.
After tax income Public Good U0 U1 D C E
SLIDE 25
ECON 551: Lecture 10 25 of 40
Matching Grant
So, imagine the feds care about the level of public goods spending at the provincial level. We have seen that block grants might not lead to the desired level of public goods spending. Is there another way? What if the federal government offers a ‘matching grant’? This kind of grant offers a match for every dollar spent at the subnational level. For example, from 1966 to 1996, the Canada Assistance Plan was a federal transfer that paid $1 for every dollar spent by provinces on social assistance (welfare) programs. Essentially, the provinces were spending 50 cent dollars. Here’s what it looks like:
SLIDE 26
ECON 551: Lecture 10 26 of 40
Matching Grant
In this diagram we see: Initial equilibrium at F. Matching grant is like a price change, has no effect at spending=0, but flattens the budget line as spending increases. Unless the public good is a giffen good, spending will increase.
After tax income Public Good U0 U1 G F
SLIDE 27
ECON 551: Lecture 10 27 of 40
What can go wrong here?
If the feds want to get the right level of the PG, they have to choose the right match rate. This might be hard. The feds don’t have any cost control—they just write cheques for however much the provinces want to spend. Both of these factors came into play in the 1990s, with the cap on CAP, and then its cessation with the start of the CHST.
SLIDE 28
ECON 551: Lecture 10 28 of 40
Agenda
- 1. Overview
- 2. Models of Federalism: Breton + Oates.
- 3. Intergovernmental Grants
- 4. Equalization Grants
SLIDE 29
ECON 551: Lecture 10 29 of 40
Equalization grants
Equalization grants are common in many countries across the world. Australia, Belgium, Spain, Germany, Canada, South Africa, Japan. The United States is a bit unique in not having a large, dedicated, equalization scheme. What is the goal of equalization grants? James Buchanan (1950): Argued that citizens of a country ought to be treated equally no matter where they live. His argument was one of horizontal equity.
SLIDE 30
ECON 551: Lecture 10 30 of 40
Why would a federal system lead to violations of horizontal equity?
Imagine a country with subnational jurisdictions constitutionally mandated to provide certain services. Say there is some resource that could be taxed locally or federally. If local taxation, then there will be incentive to move to get access to that government spending. If central taxation, no incentive to move. Therefore, residency influenced by constitution in absence of equalization, not neutral. Boadway (2004) notes substantial value judgement: For this model, we have to think the resource is not ‘owned’ by the province.
SLIDE 31
ECON 551: Lecture 10 31 of 40
What about rich and poor?
Also note that we have not at all mentioned vertical equity. The goal of equalization grants is not about ‘poor’ and ‘rich’ people and trying to tax according to the ability to pay. Neither is it about ‘poor’ and ‘rich’ regions necessarily—we care about people, not regions. Instead, equalization helps to forestall excessively high rates of taxation in resource poor regions. In essence, this means saving the immobile rich in poor regions from having to pay high taxes to fund a given level of services.
SLIDE 32
ECON 551: Lecture 10 32 of 40
The Net Fiscal Benefit and Efficient Migration
Buchanan (1950) described a calculation that took the value of public services received and subtracted the taxes paid. He referred to this difference as the fiscal residuum. This is now referred to as the net fiscal benefit or NFB. NFB = value of public services – tax burden The goal of equalization grants, is to eliminate differences in NFBs across jurisdictions. And thus, ensure fiscal neutrality for residency decisions.
SLIDE 33
ECON 551: Lecture 10 33 of 40
Anthony Scott vs. James Buchanan
Anthony Scott (1950) argued that equalization grants induce inefficient migration because it encourages people to stay in resource-poor areas. His argument is often heard so, let’s look at it: This objection is that such transfers provide amenities to poor people in resource-poor states, a situation which may be undesirable in the long-run for the following reason: The maximum income for the whole country, and so the highest average personal income, are to be achieved
- nly by maximizing national production. This in turn can be achieved only when resources and
labour are combined in such a way that the marginal product of similar units of labour is the same in all places. Buchanan (1952) rebuts this argument by arguing: For resource effects in the geographical sense to be present; ie for an income transfer to be resource distorting, like units of resource must be treated differently in different geographical areas as a result of the transfer. Therefore, the argument applies only to transfers among regions which involve differential fiscal treatment among individual ‘equals’ or among ‘like resource units’.
SLIDE 34
ECON 551: Lecture 10 34 of 40
Anthony Scott vs. James Buchanan
Scott (1952) responded by arguing that a person’s labour productivity varies by location and that they should be encouraged to leave the low productivity zones. This is a Pigouvian/paternalistic type of argument [I think.], since in the absence of government they were happy where they were. As a coda to this debate, in December 2001, Buchanan was at a conference in Montreal and was quoted in the National Post as saying: The government's equalization system may have brought benefits over the past 40 years, he said, but "it's time to wean the Atlantic area off transfer payments and make the receiving provinces and their taxpayers face their full responsibilities and spend less of the richer provinces' money."
SLIDE 35
ECON 551: Lecture 10 35 of 40
Canada’s Equalization Grant
(See Tombe CTJ 2018 for details) The Constitution Act of 1982 mandates the federal government to provide some form of equalization in Section 36, Part (2): Parliament and the government of Canada are committed to the principle of making equalization payments to ensure that provincial governments have sufficient revenues to provide reasonably comparable levels of public services at reasonably comparable levels of taxation. The formula has the following features: It takes 5 separate tax bases at the provincial level: PIT, CIT, Consumption, 50% resource revenues and property+misc. 10 prov standard. From 1982-2004 it was 5-province. (BC, SK, MB, ON, QC) If the formula delivers a positive number—you get a cheque. If negative, you get nothing. For this reason, it is called a ‘gross’ system, rather than ‘net’. NL and NS negotiated offsets for its resource revenues. 2009 a total spending cap was put on the program.
SLIDE 36
ECON 551: Lecture 10 36 of 40
Canada’s Equalization Formula: Stylized version
𝐹𝑞 = max (0, [𝑢 (𝑌𝑇 𝑂𝑇 − 𝑌𝑄 𝑂𝑄 ) 𝑂𝑄]) Where: Ep entitlement of province p. t Average tax rate, calculated as total national revenue / total national base. Xs, Xp Tax base in the provinces in the standard, and in province p. Ns,Np population in provinces in the standard, and in province p. How to interpret this formula? It compares a province’s per capita tax base to that of the standard. This is called the ‘representative tax system’ or RTS method. It refunds to the province an amount that covers the gap between the standard and itself (t x N). If the province were to charge the national average tax rate, it would therefore have enough revenue to fund expenditures at the ten-province standard.
SLIDE 37
ECON 551: Lecture 10 37 of 40
Some comments about the formula:
Why did we ever use just five provinces? In 1982 this change was made. From most accounts, it was simply a cost-saving measure. Including Alberta in the calculation would increase the averages so much that the federal government would have to pay out a lot more.
- The end result is something far short of full equalization.
Why use the tax base and not revenues? Because you don’t want the provincial government’s choices to affect its allocation.
- Problem: Smart 1998 explores the effect of elastic tax bases—provinces might choose
‘too high’ tax rates in order to shrink the base and therefore get larger equalization cheques.
- What incentive does this give provinces to increase their tax base—e.g. offshore
revenues in NL and NS. How to measure X? Different provinces have entirely different approaches to taxation. So, some standard is needed, but is at times arbitrary and arcane. It doesn’t necessarily relate to actual base definitions used in practice.
SLIDE 38
ECON 551: Lecture 10 38 of 40
How this looked in 2004-05:
SLIDE 39
ECON 551: Lecture 10 39 of 40
How this looked in 2018-19:
SLIDE 40