How Do Public Elementary Schools Affect the Value of Your Home A Quantitative Analysis of West Hartf

1
How Do Public Elementary Schools Affect the
Value of Your Home?A Quantitative Analysis of
West Hartford, CT, 1996-2005

• Research presentation by
• Jeffrey Harrelson, Laura Maloney, Drew Murphy,
• Michael Snow, and Russell Smith (Trinity College
  Class of 2007)
• Trinity Faculty Sponsors
• Professor Jack Dougherty (Educational Studies)
  and
• Professor Diane Zannoni (Economics)

2
Four Analyses

• Part 1 Boundary Type and Variable Distance
  Analysis
• Part 2 Time Period Analysis (1996-2000 vs.
  2001-2005)
• Part 3 School Racial Composition Analysis
• Part 4 Middle/High School Attendance Zone
  Analysis

3
Question How Do Public Elementary Schools Affect
the Value of Your Home?

Public School Attendance Boundary
East School Attendance Area
West School Attendance Area
B
A
Study uses econometric analysis
4
Question How Do Public Elementary Schools Affect
the Value of Your Home?
Public School Attendance Boundary
East School Attendance Area
West School Attendance Area
B
A
Study uses econometric analysis to control
for -- characteristics of homes
5
Question How Do Public Elementary Schools Affect
the Value of Your Home?
Public School Attendance Boundary
West School Attendance Area
East School Attendance Area
B
A
Study uses econometric analysis to control
for -- characteristics of homes -- neighborhood
effects
6
Question How Do Public Elementary Schools Affect
the Value of Your Home?
Public School Attendance Boundary
West School Attendance Area
East School Attendance Area
B
A
Study uses econometric analysis to control
for -- characteristics of homes -- neighborhood
effects -- school quality characteristics
7

• Study Specifications
• Area Studied
• -- West Hartford, CT
• School Quality
• House
• Neighborhood

-- Twenty Eight Different Boundaries
8
West Hartfords 26 of 28 School Attendance Borders
9
Example of Attendance Boundary Morley-Whiting
Lane (Border Code 120)
10
Example of Attendance Boundary Morley-Whiting
Lane (Border Code 120) Green Dots represent
house sold during sample period. White Dots
represent other homes.
11
HOUSE PRICE f (school quality, house
characteristics, neighborhood)

• LNHOUSEP The logarithm of the price of a house
  deflated to year 2000 dollars using an index of
  the average sales price of houses in West
  Hartford.

12
HOUSE PRICE f (school quality, house
characteristics, neighborhood)

• TEST
• The percentage of 4th grade students scoring at
  goal level on the Connecticut Mastery Test (CMT)
  averaged across three separate tests (math,
  reading, and writing).
• This measurement resembled how CMT results were
  reported in the local newspaper graphics.

13
HOUSE PRICE f (school quality, house
characteristics, neighborhood)
Hartford Courant Jan 6, 1999
14
HOUSE PRICE f (school quality, house
characteristics, neighborhood)

• BED The number of bedrooms in the house.
• BATH  The number of bathrooms in the house.
• LOTSIZE  The size of the property (in square
  feet).
• INTSF  The internal square footage of the house
  (in square feet).

15
HOUSE PRICE f (school quality, house
characteristics, neighborhood)

• BOUND Dummy variables for each of the
  twenty-six elementary school borders between two
  attendance areas. (25 Dummy Variables).

16
The Data

• In order to create a data sample, geographical
  information system (GIS) computer mapping
  techniques were used to spatially link two types
  of data elementary schools and single-family
  homes.
• During the period of study, the West Hartford
  Public Schools district consisted of 11
  elementary schools, whose residential attendance
  areas come together to form 28 borders

17

• The final sample included 8,736 qualified
  single-family home sales transactions from
  1996-2005, each linked to one of 11 elementary
  schools and one of 26 school attendance borders.

18
Descriptive Statistics of Sample
19
Part 1 Boundary Type and Variable Distance
Analysis

• Replicating Blacks model, Part 1 examined how
  the relationship between school test and home
  price changed, using two different ways to
  control for neighborhood effects

• Boundary Type
• Variable Distance

20

• Four Restrictions
• Natural Boundaries
• Transportation Set A
• Transportation Set B
• Transportation Set C

21
Map of Attendance School Boundary Shows houses
0.15 miles from the school attendance boundary
(between Morley and Whiting Lane Elementary).
Three variable distances -0.15 miles from the
boundary -0.25 miles from the boundary -0.35
miles from the boundary
22
Map of Attendance School Boundary 0.25 miles
from the school attendance boundary
23
Map of Attendance School Boundary 0.35 miles
from the school attendance boundary
24
Table 1Coefficient of TEST Score on House Price
for all Geographical Restrictions over
Time(t-statistic in parenthesis)Dependent
Variable ln(HOUSEP)
25
Table 2 Increase in Housing Price for Various
Percentage Point Gain in Scores with Most
Restrictive Geography
one standard deviation
26
Part 2 Time Period Analysis(1996-2000 vs.
2001-2005)

• What effect did increasing levels of access to
  school test data have on home prices?
• Expanding upon Blacks model, this study divided
  the 10-year time period (1996-2005) in half to
  measure any change in the test-price relationship
  over time.
• Reasons for choosing 1996-2005
• -West Hartfords recent redistricting in 1995
  which held through 2005
• -CMT test generations and scale were comparable
  over period
• -Expansion of school accountability policies, as
  well as the internet boom.

27
Table 3 Coefficient of TEST Score on House Price
for all Geographical Restrictions over
Time(t-statistic in parenthesis)Dependent
Variable ln(HOUSEP)
28
Table 4 Increase in Housing Price for Various
Percentage Point Gain in Test Scores with Most
Restrictive Geography
one standard deviation
29
Thanks

• Special thanks to Dave Tatem, Professor Jack
  Dougherty, and Professor Diane Zannoni
• Ryan Butler 08, Molly Stumbras 07, and Ben
  Willig 08
• Thanks to Sandra Black for making a special
  appearance