Economics 147

1
Economics 147

• Winter 2005
• Lecture 8

2
Outline for today

• The Tiebout model and traditional school choice
• How it fits into the course
• Arguments for and against centralization of
  education
• The model
• Assumptions
• Policy implications

3
Where are we in the course?

• Started with human capital investment problem
• Expand it to allow possible return to schooling
  from signaling
• Allow school quality to matter in determining
  return to education
• Education production functions
• Public vs. private
• Class size, peers

4
Theme throughout

• Must identify causal effects of schooling (or
  particular quality component)
• Remember that people choose quantity/quality of
  schooling
• These choices may be influenced by omitted
  variables

5
Economic framework

• Optimal investment decision demand for
  education
• Education production function supply of
  education
• Next markets for education

6
What level of government should provide education?

• Why should government provide education at all?
• Credit market failures
• Externalities
• Given government provision, how centralized
  should it be?
• Most centralized ? most
  decentralized
• Federal ? State ? County ?
  Community

7
(No Transcript)
8
Unfunded mandates

• Do not show up in intergovernmental revenues and
  expenditures
• 1970s federal deficit ? federal govt passed on
  responsibilities to the states
• Late 1980s, states entered fiscal crises, passed
  on responsibilities to localities

9
Arguments for centralized provision of government
services

• Efficiency arguments
• Economies of scale
• Spillovers across jurisdictions
• Equity arguments
• Greater capacity to redistribute

10
Arguments against centralized provision of
government services

• Efficiency arguments
• Match individual preferences more closely
• Promote competition in provision of the good
• Matters if you are worried about bureaucrats
  trying to maximize their budgets
• Diseconomies of scale
• Equity arguments
• Hard to make

11
Tiebout (1956) model of local provision

• Starting point people understate their
  preferences for government spending (G) if they
  will be taxed accordingly (free-rider problem)
• Tiebouts insight dont ask people how much G
  want, make them reveal it through where they
  choose to live
• Different communities offer different levels of G
  and different tax rates
• Individuals vote with their feet
• Eliminates ability to free ride ? improves
  efficiency

12
Tiebout hypothesis

• Public goods provision will be efficient at the
  local level (under many assumptions)

13
Model

• Many individuals divide themselves into different
  local jurisdictions (towns)
• Each town has N residents
• Private good x (px1)
• Public good G (pG1)
• Within a town, residents are identical
• Indiv preferences given by U(x,G)
• Indiv budget constraint
• x G/N y

14
Tiebout makes strong assumptions

• Need a competitive market
• Consumers are fully mobile
• no moving costs
• Consumers are fully informed
• perfect information on govt services and taxes
• Unlimited number of communities exist
• at least as many as types of people

15
Assumptions about production function

• No externalities or spillovers across communities
• Positive externalities make local residents
  supply G at an inefficiently low level
• Negative externalities (like pollution) make
  local residents supply G at an inefficiently high
  level
• Constant marginal costs

16
Assumptions about financing

• Need these assumptions to prevent free riding
• Must do one of these
• Assume a head tax (same level of tax for
  everyone)
• Assume a property tax based on home value and
  assume zoning

17
Are competitive market assumptions realistic?

• Consumers are fully mobile
• US Census 16 of Americans have a different
  address in a given year than they did the year
  before
• Moving costs have declined over time but are
  still sizable, especially non-monetary costs
• Consumers have perfect information on govt
  services and taxes
• Blinder Krueger evidence on federal level next
• At least as many communities as types of
  individuals
• Varies by metropolitan area
• One school district per 36,025 people in CA
• One school district per 14,302 people in NJ
• One school district per 254,016 people in FL

18
(No Transcript)
19
Source Blinder Krueger
20
Are production function assumptions realistic?

• No externalities or spillovers across communities
• Low level of environmental regulation/enforcement
  in Mexico affects Imperial Beach (CA) water
  quality
• State-subsidized UC education benefits other
  states
• Constant marginal costs
• Depends on the good being provided

21
Are financing assumptions realistic?

• Must do one of these
• Assume a head tax (same level of tax for
  everyone)
• - unrealistic
• Assume a property tax based on home value and
  assume zoning
• - we see this in reality

22
Why zoning?

• Property tax is applied to the value of the house
• You can free ride by buying a small house in a
  rich neighborhood
• You benefit from your neighbors high tax
  payments on fancy houses but pay little yourself
• Rich want to keep the poor out of their
  neighborhoods to prevent this free riding
• Zoning is a way to do this

23
Example Lincoln, MA original zoning law
24
Characteristics of (our stylized) equilibrium

• HHs set MB of school quality MC
• Households sort themselves into communities by
  preference for school quality
• House prices capitalize differences in school
  quality
• Producers of schooling produce efficiently