1# RECORD a2a7539a9c4a1cf05650ab56eedb83ae824bb6362# TITLE: Nonlinear Effects of School Quality on House Prices3# YEAR: 2003.0 VENUE: nan DOI: 10.20955/r.92.185-2044# AUTHORS: Abbigail J. Chiodo; Rubén Hernández-Murillo; Michael T. Owyang5# PAGES: 29 CANDIDATE TABLE PAGES (1-based): [25, 27, 28]6# ABSTRACT: We reexamine the relationship between quality of public schools and house prices and find it to be nonlinear. Unlike most studies in the literature, we find that the price premium parents must pay to buy a house in an area associated with a better school increases as school quality increases. This is true even after controlling for neighborhood characteristics, such as the racial composition of neighborhoods, which is also capitalized into house prices. In contrast to previous studies that use the boundary discontinuity approach, we find that the price premium from school quality remains substantially large, particularly for neighborhoods associated with high-quality schools.789===== PAGE 1 (score 3, numeric-paren 0, table-words 0, var-words 3) =====1011Nonlinear Effects of School Quality on House Prices12ECONOMIC RESEARCH13FEDERAL RESERVE BANK OF ST. LOUIS14WORKING PAPER SERIES15Authors16Abbigail J. Chiodo, Rubén Hernández-Murillo, and Michael T. Owyang17Working Paper Number182003-039E19Revision Date20September 200921Citable Link22https://doi.org/10.20955/wp.2003.03923Suggested Citation24Chiodo, A.J., Hernandez-Murillo, R., Owyang, M.T., 2009; Nonlinear Effects of School25Quality on House Prices, Federal Reserve Bank of St. Louis Working Paper 2003-26039. URL https://doi.org/10.20955/wp.2003.03927Published In28Federal Reserve Bank of St. Louis Review29Publisher Link30https://doi.org/10.20955/r.92.185-20431Federal Reserve Bank of St. Louis, Research Division, P.O. Box 442, St. Louis, MO 6316632The views expressed in this paper are those of the author(s) and do not necessarily reflect the views of the Federal Reserve33System, the Board of Governors, or the regional Federal Reserve Banks. Federal Reserve Bank of St. Louis Working Papers34are preliminary materials circulated to stimulate discussion and critical comment.353637===== PAGE 25 (score 92, numeric-paren 61, table-words 6, var-words 9) =====3839Table 3: Education Regressions: Full Sample.40(1)41(2)42(3)43log adj price44log adj price45log adj price46test math470.21734***480.22192***490.31693***50(7.79)51(7.13)52(7.70)53test math sq540.03002550.0155556(1.48)57(0.76)58test math cb59-0.03606**60(-2.60)61house bedrooms620.01062630.01502640.0157565(1.09)66(1.52)67(1.62)68house bathrooms690.14086***700.14413***710.13458***72(4.75)73(4.93)74(4.44)75house baths sq76-0.0061277-0.0074078-0.0050179(-1.14)80(-1.37)81(-0.89)82house age830.00065840.00057850.0012386(0.37)87(0.31)88(0.67)89house age sq900.00002910.00002920.0000293(1.35)94(1.31)95(1.03)96house lotarea970.00123***980.00120***990.00119***100(4.21)101(4.27)102(4.17)103house livingarea1040.45365***1050.44475***1060.43526***107(20.02)108(17.35)109(19.05)110house stories1110.39693***1120.38775***1130.37835***114(11.29)115(10.58)116(10.87)117house rooms1180.07484***1190.07421***1200.07245***121(10.10)122(10.21)123(10.11)124cenbl fem125-0.00061126-0.00050127-0.00053128(-0.88)129(-0.73)130(-0.79)131cenbl nonwhite132-0.00221***133-0.00277***134-0.00257***135(-3.62)136(-5.06)137(-4.57)138cenbl p5 14139-0.00017140-0.00033141-0.00021142(-0.19)143(-0.38)144(-0.24)145tax146-0.04636147-0.04457148-0.03562149(-1.65)150(-1.51)151(-1.28)152cons15310.00143***1549.99065***1559.96337***156(59.89)157(57.55)158(58.13)159N16038,65616138,65616238,656163R21640.6971650.6991660.702167Adjusted R21680.6971690.6981700.702171t-Statistics in parentheses.172* Significant at 10%; ** significant at 5%; *** significant at 1%.17324174175176===== PAGE 26 (score 7, numeric-paren 5, table-words 1, var-words 0) =====177178Table 4: Specification Tests: Full Sample179Premium Model f(Y ) = ψ1Y + ψ2Y 2 + ψ3Y 3180Linear181Quadratic182Cubic183With Neighborhood Controls184Null hypothesis185ψ1 = 0186ψ1 = ψ2 = 0187ψ1 = ψ2 = ψ3 = 0188Wald F-Statistic18960.757***19027.686***19130.665***192Null hypothesis193ψ2 = 0194ψ2 = ψ3 = 0195Wald F-Statistic1962.1921977.446***198Null hypothesis199ψ3 = 0200Wald F-Statistic2016.754**202* Significant at 10%; ** significant at 5%; *** significant at 1%20325204205206===== PAGE 27 (score 184, numeric-paren 139, table-words 9, var-words 9) =====207208Table 5: Education Regressions: Boundary Sample209(1)210(2)211(3)212(4)213(5)214(6)215log adj price216log adj price217log adj price218log adj price219log adj price220log adj price221test math2220.06437**2230.06274***2240.046592250.03227*2260.03579*2270.03172228(2.58)229(2.90)230(1.64)231(1.78)232(1.93)233(1.20)234test math sq2350.02656**2360.02909**2370.02209**2380.02284**239(2.47)240(2.47)241(2.48)242(2.40)243test math cb2440.005142450.00137246(0.73)247(0.21)248house bedrooms2490.03726***2500.03730***2510.03749***2520.03816***2530.03805***2540.03809***255(3.88)256(3.89)257(3.90)258(4.02)259(4.01)260(4.00)261house bathrooms2620.10834***2630.10785***2640.10792***2650.10349***2660.10318***2670.10320***268(5.78)269(5.80)270(5.82)271(5.81)272(5.82)273(5.83)274house baths sq275-0.00529*276-0.00533*277-0.00535*278-0.00488279-0.00491280-0.00491281(-1.68)282(-1.70)283(-1.71)284(-1.58)285(-1.60)286(-1.60)287house age288-0.00408***289-0.00411***290-0.00412***291-0.00453***292-0.00454***293-0.00454***294(-2.73)295(-2.75)296(-2.76)297(-3.11)298(-3.13)299(-3.14)300house age sq3010.00004***3020.00004***3030.00004***3040.00004***3050.00004***3060.00004***307(2.89)308(2.91)309(2.92)310(3.15)311(3.16)312(3.17)313house lotarea3140.00089**3150.00089**3160.00089**3170.00088**3180.00088**3190.00088**320(2.41)321(2.41)322(2.41)323(2.39)324(2.40)325(2.39)326house livingarea3270.35315***3280.35228***3290.35236***3300.34332***3310.34297***3320.34301***333(15.43)334(15.29)335(15.29)336(15.52)337(15.49)338(15.49)339house stories3400.27574***3410.27559***3420.27558***3430.26621***3440.26625***3450.26626***346(9.30)347(9.30)348(9.31)349(9.55)350(9.57)351(9.57)352house rooms3530.05974***3540.05952***3550.05945***3560.05902***3570.05893***3580.05891***359(7.38)360(7.33)361(7.31)362(7.43)363(7.40)364(7.39)365cenbl fem366-0.00044367-0.00039368-0.00039369(-0.66)370(-0.59)371(-0.59)372cenbl nonwhite373-0.00219***374-0.00223***375-0.00222***376(-3.50)377(-3.56)378(-3.55)379cenbl p5 143800.00154**3810.00153**3820.00154**383(2.25)384(2.24)385(2.25)386tax387-0.06787***388-0.05526***389-0.05465***390(-3.21)391(-2.88)392(-2.73)393cons39411.13260***39511.12998***39611.13935***3978.86314***3988.72871***3998.72454***400(32.85)401(32.96)402(32.97)403(59.31)404(62.61)405(60.54)406N407101904081019040910190410101824111018241210182413R24140.7694150.7704160.7704170.7724180.7724190.772420Adjusted R24210.7634220.7634230.7634240.7664250.7664260.766427Boundary fixed effects428Yes429Yes430Yes431Yes432Yes433Yes434t-Statistics in parentheses.435* Significant at 10%; ** significant at 5%; *** significant at 1%.43626437438439===== PAGE 28 (score 14, numeric-paren 11, table-words 1, var-words 0) =====440441Table 6: Specification Tests: Boundary Sample442Premium Model f(Y ) = ψ1Y + ψ2Y 2 + ψ3Y 3443Linear444Quadratic445Cubic446Without Neighborhood Controls447Null hypothesis448ψ1 = 0449ψ1 = ψ2 = 0450ψ1 = ψ2 = ψ3 = 0451Wald F-Statistic4526.632**4534.658**4543.130**455Null hypothesis456ψ2 = 0457ψ2 = ψ3 = 0458Wald F-Statistic4596.115**4603.114**461Null hypothesis462ψ3 = 0463Wald F-Statistic4640.527465With Neighborhood Controls466Null hypothesis467ψ1 = 0468ψ1 = ψ2 = 0469ψ1 = ψ2 = ψ3 = 0470Wald F-Statistic4713.178*4723.581**4732.381*474Null hypothesis475ψ2 = 0476ψ2 = ψ3 = 0477Wald F-Statistic4786.166**4793.102**480Null hypothesis481ψ3 = 0482Wald F-Statistic4830.043484* Significant at 10%; ** significant at 5%; *** significant at 1%48527486487488===== PAGE 29 (score 9, numeric-paren 0, table-words 2, var-words 1) =====489490Table 7: Implied House Premia from Math Test Scores491Change in Math score (SD)4920.5493Equivalent percentage of mean4944.6%495Regression model496Full sample with497Boundary sample with498Boundary sample with499neighborhood controls500no neighborhood controls501neighborhood controls502Linear coefficient5030.217345040.221925050.316935060.064375070.062745080.046595090.032275100.035795110.03172512Quadratic coefficient513–5140.030025150.01555516–5170.026565180.02909519–5200.022095210.02284522Cubic coefficient523–524–525-0.03606526–527–5280.00514529–530–5310.00137532Case 1 (Mean score - 1 SD)533Percent increase in house price53410.875358.8453611.535373.225381.155390.605401.615410.13542-0.01543Dollar value at mean (full sample)54416,09254513,09754617,0665474,7665481,6965498855502,389551197552-11553Dollar value at mean (boundary sample)55415,43555512,56255616,3695574,5715581,6265598495602,292561189562-10563Case 2 (Mean score)564Percent increase in house price56510.8756611.8556715.785683.225693.805703.125711.615722.345732.17574Dollar value at mean (full sample)57516,09257617,54257723,3745784,7665795,6295804,6225812,3895823,4685833,219584Dollar value at mean (boundary sample)58515,43558616,82658722,4195884,5715895,3995904,4335912,2925923,3265933,088594Case 2 (Mean score + 1 SD)595Percent increase in house price59610.8759714.855989.235993.226006.466017.196021.616034.556044.77605Dollar value at mean (full sample)60616,09260721,98860813,6626094,7666109,56261110,6426122,3896136,7396147,058615Dollar value at mean (boundary sample)61615,43561721,09061813,1046194,5716209,17162110,2076222,2926236,4646246,770625The premium is computed from the logarithm specification ∆p/p = ∆ln(p) = ∆f(µ), so the percent change626in house prices is given by ∆f(µ) = f(µ1) −f(µ0) and the premium at the mean price is ∆f(µ) × ¯p.62728628