GPP 501 Microeconomic Analysis for Public Policy Fall 2017 Given - - PowerPoint PPT Presentation

GPP501: Lecture Nov. 8th 1 of 26

GPP 501 Microeconomic Analysis for Public Policy Fall 2017

Given by Kevin Milligan Vancouver School of Economics University of British Columbia Lecture November 8th Public Goods: Local Public goods

GPP501: Lecture Nov. 8th 2 of 26

Agenda

1. Club Goods 2. Local Public Goods: The Tiebout model 3. Multi-level government: federalism

GPP501: Lecture Nov. 8th 3 of 26

Club goods

James Buchanan, 1986 Nobel Prize winner

We have studied pure private goods and pure public goods.  But for many public goods we considered them non-rival only up to some point of congestion. Buchanan wrote the paper “An Economic Theory of Clubs” to take congestion seriously and think about how it changes public goods analysis.

GPP501: Lecture Nov. 8th 4 of 26

Club goods

Here is a definition: Definition: A club is a voluntary group of individuals who derive mutual benefit from sharing one or more of the following:  Production costs  Members’ characteristics (mutual benefits)  A good with excludable benefits (singing in a choir) Any good provided by the club must be  Excludable, so that non-members don’t get the benefit.  Rejectable, so that membership is voluntary.

GPP501: Lecture Nov. 8th 5 of 26

Club goods

The number of people in the club: s The per-capita monetary costs are: C(s) Q: Why is this sloping down? Q: How are fixed and variable costs important for club goods?

s C(s) Total Cost Per capita

GPP501: Lecture Nov. 8th 6 of 26

Club goods

The number of people in the club: s The per-capita benefits are: C(s) Q: Why do we assume the benefit curve has this shape? Q: What role does congestion play here?

s B(s ) Total Benefit Per capita

GPP501: Lecture Nov. 8th 7 of 26

Club goods

The goal is to maximize the social surplus—find the level of s that has the maximum Benefit-Cost gap.  Point s* looks interesting. Q: What happens to costs and benefits if you add a little s past s*?

s C(s) B(s) s* Total Benefit ~ Total Cost Per capita

GPP501: Lecture Nov. 8th 8 of 26

Club goods: summary

So, to summarize the circumstances where club goods model fits:  Goods with high fixed costs  Some degree of non-rival consumption.  Congestion arises after some relevant threshold.

GPP501: Lecture Nov. 8th 9 of 26

Club goods: Applications

 This is applicable to many situations: road tolls, military alliances that share resources, transit systems.  What does this say about public good provision by centralized government? If PGs are excludable, then little groups will form to provide nonrival public

• goods. This limits the role of government.

 Is the club replicable? If so, then the whole population can form clubs of size s*, and everyone is happy. Local swimming pools.

GPP501: Lecture Nov. 8th 10 of 26

Agenda

1. Club Goods 2. Local Public Goods: The Tiebout model 3. Multi-level government: federalism

GPP501: Lecture Nov. 8th 11 of 26

Local Public Goods

The Samuelson framework applies to public goods that provide services to all within a jurisdiction.  But what if the benefits to a public good are local? For example, a park or a community garden.  Does the ‘localness’ of a public good change anything? The ‘Club Goods’ model focuses on per capita costs and benefits.  Assumes everyone is the same.  What if people are heterogeneous?

GPP501: Lecture Nov. 8th 12 of 26

The Tiebout Model

“A pure theory of local expenditures,” Charles M. Tiebout (JPE 1956) In a world with:  Many jurisdictions  Public goods that provide local benefits  Heterogeneous citizens. Charles Tiebout argued that:  Local governments would compete for citizens like firms compete for customers.  Heterogeneous individuals would be best served by ample community choice, rather than by public good provision by some distant central government.

GPP501: Lecture Nov. 8th 13 of 26

The Tiebout Model

“Vote with your feet” The mechanism works through mobility.  Families are assumed to choose a community that provides a tax-service balance that fits best their needs.  This provides a decentralized solution. The model has been extraordinarily influential, with vast applications across study of local government, schools, local taxation, real estate economics, and more generally urban economics. It has also spawned a great deal of more purely theoretical work. Let’s look a little more closely…

GPP501: Lecture Nov. 8th 14 of 26

The Tiebout Model: Assumptions

• 1. Consumers are perfectly mobile—no cost of moving.
• 2. Consumers have perfect information about taxes and public goods in

different communities.

• 3. Consumers are endowed with income; no labour market.
• 4. The number of communities is large with a wide variety of goods and tax

mixes.

• 5. There is no benefit of the public good external to the community; exclusion

is possible.

• 6. Communities’ objective is to reach the optimal size by attracting or getting

rid of consumers. Decisions are made by a ‘city manager’.

GPP501: Lecture Nov. 8th 15 of 26

The Tiebout Model: Outcomes

The predicted outcomes in this world are:

• 1. Optimal allocation of public goods within communities.
• 2. All communities will be at optimal size (defined by minimum efficient

scale).

• 3. Within communities: homogenous in preference for public goods and

taxes.

• 4. Across communities: heterogeneous choices on public goods and taxes.

GPP501: Lecture Nov. 8th 16 of 26

The Tiebout Model: Comments

Comments Not clear how the results carry through when assumptions relaxed; not proven by Tiebout for a more general case. Prediction: Are communities really homogenous? This gives the model empirical content that can be tested. What if real estate was in the model?  Imagine that two houses are entirely identical, but one is in town 1 and the

• ther in town 2.

 Imagine that the taxes in town 1 are zero, but the annual taxes in town 2 are t.  Assume that no valued services result from the taxes.  In equilibrium, the (annualized) price of the houses should satisfy: p1 = p2(1- t). That is, the tax would be capitalized into the price of houses in town 2.

GPP501: Lecture Nov. 8th 17 of 26

Agenda

1. Club Goods 2. Local Public Goods: The Tiebout model 3. Multi-level government: federalism

GPP501: Lecture Nov. 8th 18 of 26

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 NewYork 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.”

GPP501: Lecture Nov. 8th 19 of 26

In 1926, the shares were: 37.8 fed, 20.2 prov, 42.0 local. (Source: Musgrave, Musgrave, and Bird)

GPP501: Lecture Nov. 8th 20 of 26

GPP501: Lecture Nov. 8th 21 of 26

GPP501: Lecture Nov. 8th 22 of 26

Breton (1965): “A Theory of Government Grants”

https://www.jstor.org/stable/140062 Theory hinges on the provision of public goods.  Which level of government should provide them? Breton describes public goods with various degrees

• f ‘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)

GPP501: Lecture Nov. 8th 23 of 26

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

• f 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.

GPP501: Lecture Nov. 8th 24 of 26

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 others’ threat points, free-riding by jurisdictions, enforceability of agreements—governments might renege.

GPP501: Lecture Nov. 8th 25 of 26

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.”

GPP501: Lecture Nov. 8th 26 of 26

The Oates ‘Decentralization Theorem.’

“If there are no spillovers, it is always at least as good for public goods to be provided by local governments” 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