Summary EC325 Exam Relevant Class Reading Notes

LECTURE 3 – EXTERNALITIES
Does Air Quality Matter? Evidence from the Housing Market [Chay & Greenstone, 2005]

OBJECTIVE:
Estimate the willingness to pay for changes in air quality by looking at how house prices change in response to
changes in air quality driven by the Clean Air Act

Clean Air Act Amendments (1970):
Legislation imposing strict regulations on polluters in ‘nonattainment’ counties, defined by concentrations of TSPs
that exceed a federally set ceiling
 Nonattainment counties have more stringent restrictions on polluters than attainment counties

WHY FOCUS ON TSPs?
 It had the most harmful effects of all pollutants regulated by CAAAs
 Including other air pollutants would have severely restricted the sample size of the data

METHOD:
Hedonic price method was used to estimate the economic value of clear air (a non­market amenity) to individuals
 It predicted that air pollution will be negatively correlated with housing prices (all other characteristics held
constant)
CAAA approach = non­attainment status used as an IV for ΔTSPs in the first differenced equations for
Δ house prices

BASIC FRAMEWORK:
 House Price a ,t =β 0+ β1 Amenity a ,t + X a ,tγ + ε a , t
ΔHouse Price
 Willingness¿ Pay=β 1=
ΔAmenity
o Gives us elasticity of house prices wrt pollution

RESULTS:
 Regulations in nonattainment counties were successfully enforced following the CAAAs
 HEDONIC APPROACH: TSPs levels only led to a small property values
o Resulting in small marginal willingness to pay for environmental amenities
o β 1=¿ ­0.04 to ­0.07
o This was due to OVB & heterogeneity across individuals tastes for clean air
 Leads to individuals self­selecting into locations
 Self­selection bias was small relative to influence of OVB
 Improvements in air quality induced by TSPs nonattainment designations can be associated with a $45bn
aggregate increase in housing values in nonattainment counties
o BUT net welfare effect is unknown since reliable estimates of social costs/regulations are
unavailable
 CAAA APPROACH: Individuals place a higher value on clean air than previously recognised
o Pollution level by 1% = price levels by 36 percentage points
 Overall quite large magnitude
 PROBLEM: we’re assuming people perfectly perceive the changes in amenities that
they’re exposed to

RELIABILITY OF THE DATA USED:
 TSPs attainment/nonattainment designations:
o Reasonable approximation to EPA’s actual selection rule was used
o There could be some room for error & imprecision
 Housing values:
o CCDBs used are comprehensive, reliable & complete with info for every US county
o Census data contained fewer variables on characteristics of homes & neighbourhoods than was
ideal
o Existence of a national housing market was assumed

,  Unable to explore the degree of within­county taste heterogeneity & sorting
 Could lead to results understating individual­level dispersion in MWTP
o Income & other similar variables are generally excluded
 Direct test of validity of exclusion restriction unavailable
o Hedonic approach was original an individual level model
 Aggregation to county level may induce some biases

LECTURE 4 – PUBLIC GOODS [Gneezy, Uri and Rustichini, 2000]
A Fine is a Price

THE STUDY:
10 day­care centres were observed over 20 weeks (Israel, 1991­1992)
 In the 1st 4 weeks  only the no. parents who arrived late were observed
 Beginning of 5th week  a monetary fine was introduced to 6 day­care centres on those who arrived 10+
mins late (other 4 day­care centres served as a control group)
o The fine of NIS 10 was relatively small but not insignificant
 Beginning of 7th week  the fine was removed

METHOD:
 Control: initial 4 weeks
 Treatment group was selected randomly
o There was no significant difference in the behaviour of the test group & control group in the initial 4
weeks
o No significant trend in the test group in the initial 4 weeks
 All day­care centres were located in the same part of town with no important differences among them
 RDD  exogenous variation from average class size falling after every multiple of 40


THE RESULTS:
 The no. late­coming parents increased significantly after the fine was introduced
o Almost doubled in the level of the rate of delay
 At the end of the adjustment period after the fine was introduced (lasting 2­3 weeks)  the no. late coming
parents remained stable @ a higher rate than in the no­fine period
 After the fine was removed  no reductions occurred in the level of the rate of delay for the treatment group
 The results contradicted the deterrence hypothesis
o Deterrence hypothesis: predicts that the introduction of a penalty that leaves everything else
unchanged will reduce the occurrence of the behaviour subject to the fine

INTERPRETATION OF RESULTS:
 The results contradicted the deterrence theory for 2 possible reasons:
o A “large enough” fine would eventually reduce the behaviour
o The introduction of a fine can change the perception of the obligation to arrive on time
 ∴ The assumption that “everything else is left unchanged” didn’t hold
 Contract was previously incomplete  parents didn’t know how bad it was to be late (as
there was no price on being late)
 Intrinsic motivation for being a good person is removed
o People don’t want to seem greedy for money