# RECORD b633c49b2f5879640f919bdc54ae4390ee31d91d # TITLE: Time-Geographically Weighted Regressions and Residential Property Value Assessment # YEAR: 2019.0 VENUE: Journal of real estate finance and economics DOI: 10.1007/s11146-019-09718-8 # AUTHORS: Jeffrey P. Cohen; Cletus C. Coughlin; Jeffrey E. Zabel # PAGES: 27 CANDIDATE TABLE PAGES (1-based): [16, 18, 19, 24] # ABSTRACT: nan ===== PAGE 1 (score 2, numeric-paren 0, table-words 0, var-words 2) ===== Time-Geographically Weighted Regressions and Residential Property Value Assessment ECONOMIC RESEARCH FEDERAL RESERVE BANK OF ST. LOUIS WORKING PAPER SERIES Authors Jeffrey P. Cohen, Cletus C. Coughlin, and Jeffrey Zabel Working Paper Number 2019-005A Creation Date January 2019 Citable Link https://doi.org/10.20955/wp.2019.005 Suggested Citation Cohen, J.P., Coughlin, C.C., Zabel, J., 2019; Time-Geographically Weighted Regressions and Residential Property Value Assessment, Federal Reserve Bank of St. Louis Working Paper 2019-005. URL https://doi.org/10.20955/wp.2019.005 Federal Reserve Bank of St. Louis, Research Division, P.O. Box 442, St. Louis, MO 63166 The views expressed in this paper are those of the author(s) and do not necessarily reflect the views of the Federal Reserve System, the Board of Governors, or the regional Federal Reserve Banks. Federal Reserve Bank of St. Louis Working Papers are preliminary materials circulated to stimulate discussion and critical comment. ===== PAGE 16 (score 97, numeric-paren 72, table-words 4, var-words 9) ===== 15 Table 2: Town Fixed Effects Results Dependent Variable: logprice CT MA Log(living area sqft) 0.458*** 0.323*** (0.032) (0.010) Log(lotsize) 0.0921*** 0.063*** (0.011) (0.005) 2 bathrooms+ 0.0436*** 0.051*** (0.007) (0.003) 3 bathrooms+ 0.122*** 0.117*** (0.009) (0.008) 4 or more bathrooms 0.282*** 0.070*** (0.018) (0.012) Age>=10 and < 30 0.166*** -0.033*** (0.012) (0.006) Age>=30 and <=50 0.0588*** -0.127*** (0.008) (0.007) Age>=50 -0.0248 -0.177*** (0.013) (0.011) 2 bedrooms 0.0357** 0.010 (0.030) (0.013) 3 bedrooms 0.0500*** 0.019 (0.030) (0.014) 4 bedrooms 0.0356** 0.032** (0.028) (0.014) 5 or more bedrooms 0.0566*** 0.006 (0.026) (0.014) 5 total rooms -0.0167* 0.024*** (0.009) (0.005) 6 total rooms -0.0210 0.026*** (0.014) (0.007) 7 total rooms -0.00341 0.034*** (0.018) (0.008) 8 total rooms 0.0107 0.064*** (0.022) (0.008) 9 total rooms 0.0293 0.106*** (0.025) (0.009) 10 to 14 total rooms 0.0577** 0.130*** (0.026) (0.011) 15 or more rooms 0.196*** 0.107*** (0.050) (0.023) 1 half bathroom - 0.073*** - (0.002) 2 or more half bathrooms - 0.112*** - (0.007) ===== PAGE 17 (score 8, numeric-paren 0, table-words 1, var-words 6) ===== 16 R2 0.844 0.700 N 258,473 609,123 * p < 0.10, ** p < 0.05, *** p < 0.01, Clustered standard errors in parentheses Year dummies included + For the CT results, the data from Warren Group are not broken up into half baths and full baths; rather, 2 bathrooms include 2 or 2.5 bathrooms and 3 bathrooms includes 3 or 3.5 bathrooms ===== PAGE 18 (score 28, numeric-paren 0, table-words 4, var-words 8) ===== 17 Table 3a: Summary Statistics for Coefficient Estimates for Town-Level OLS, GWR, and TGWR: CT data, b=200 Town-Level OLS GWR TGWR Variable Mean SD Mean SD Mean SD Log(living area sqft) 0.460846 0.147036 0.494622 0.214319 0.493798 0.194698 Log(lotsize) 0.087939 0.05912 0.098415 0.086663 0.097414 0.078184 Age>=10 and < 30 0.158827 0.066766 0.138405 0.162553 0.136647 0.126485 Age>=30 and <=50 0.056267 0.052834 0.043718 0.122411 0.042734 0.087073 Age>=50 -0.03772 0.076631 -0.03105 0.205574 -0.03186 0.09519 2 bathrooms 0.034511 0.026955 0.030772 0.202676 0.030297 0.138418 3 bathrooms 0.107408 0.046833 0.111545 0.227011 0.111214 0.129238 4 or more bathrooms 0.198859 0.100188 0.190862 0.272103 0.195438 0.242935 2 bedrooms -0.00067 0.068846 0.827136 2.461339 0.199719 1.268229 3 bedrooms 0.000652 0.066883 0.830924 2.465192 0.199368 1.254315 4 bedrooms -0.0081 0.072914 0.819265 2.468122 0.190706 1.258414 5 or more bedrooms -0.03853 0.074284 0.772312 2.432396 0.160399 1.264407 5 total rooms -0.01865 0.077376 0.099089 1.074278 -0.0104 0.343364 6 total rooms -0.01302 0.08898 0.099435 1.087259 -0.01002 0.371523 7 total rooms 0.002402 0.103602 0.113502 1.107922 0.002799 0.359325 8 total rooms 0.013262 0.114643 0.122437 1.115686 0.011115 0.379564 9 total rooms 0.031008 0.124483 0.133033 1.119346 0.022335 0.391953 10-14 total rooms 0.022581 0.130131 0.124008 1.109783 0.015205 0.410715 15 or more total rooms -0.0099 0.266177 0.054643 0.829567 -0.06913 0.545862 Constant 7.692293 0.959022 6.850662 2.860154 7.611603 1.842803 Note: Town-Level regressions include year dummies ===== PAGE 19 (score 30, numeric-paren 0, table-words 4, var-words 9) ===== 18 Table 3b: Summary Statistics for Coefficient Estimates for Town-Level OLS, GWR, and TGWR: MA data, b=200 Town-Level OLS GWR TGWR Variable Mean SD Mean SD Mean SD Log(living area sqft) 0.315 0.099 0.309 0.177 0.302 0.157 Log(lotsize) 0.070 0.052 0.078 0.093 0.079 0.079 Age>=10 and < 30 -0.012 0.077 -0.008 0.452 0.019 0.226 Age>=30 and <=50 -0.102 0.094 -0.092 0.506 0.108 0.239 Age>=50 -0.153 0.120 -0.142 0.521 0.158 0.250 2 bathrooms 0.057 0.030 0.055 0.152 0.052 0.080 3 bathrooms 0.097 0.065 0.093 0.247 0.085 0.153 4 or more bathrooms 0.054 0.130 0.043 0.402 0.027 0.407 1 half bathroom 0.069 0.052 0.124 0.779 0.156 0.962 2 or more half bathrooms 0.099 0.079 0.145 0.812 0.175 0.984 2 bedrooms 0.039 0.127 1.621 3.710 0.805 2.684 3 bedrooms 0.045 0.124 1.628 3.708 0.815 2.684 4 bedrooms 0.059 0.132 1.637 3.711 0.826 2.689 5 or more bedrooms 0.028 0.137 1.557 3.666 0.786 2.683 5 total rooms 0.025 0.070 0.354 1.818 0.178 1.254 6 total rooms 0.041 0.076 0.393 1.879 0.196 1.257 7 total rooms 0.055 0.084 0.410 1.887 0.211 1.263 8 total rooms 0.077 0.089 0.433 1.890 0.235 1.270 9 total rooms 0.105 0.099 0.455 1.884 0.260 1.271 10-14 total rooms 0.109 0.114 0.441 1.856 0.252 1.254 15 or more total rooms 0.055 0.270 0.075 0.791 0.074 0.946 Constant 9.723 0.675 8.114 4.240 9.162 3.233 Note: Town-Level regressions include year dummies ===== PAGE 20 (score 6, numeric-paren 0, table-words 1, var-words 0) ===== 19 Table 4: Summary Statistics for COD and PRD b=200 variable mean sd min max CT data OLS COD 3.44 3.41 -7.43 12.36 PRD 1.04 0.05 0.94 1.14 GWR COD 5.75 1.76 2.84 12.77 PRD 1.00 0.001 1.00 1.01 GWR with Time COD 6.29 1.97 3.03 13.42 PRD 1.02 0.03 0.99 1.09 MA data OLS COD 11.89 3.47 5.60 28.02 PRD 1.11 0.03 1.06 1.27 GWR COD 6.53 2.13 1.18 14.98 PRD 1.05 0.02 1.02 1.15 GWR with Time COD 6.06 2.17 0.04 15.05 PRD 1.02 0.01 1.00 1.07 ===== PAGE 24 (score 22, numeric-paren 0, table-words 7, var-words 2) ===== 23 Table 7a: Summary Statistics for Percent Absolute Prediction Error: CT Data, TGWR Bandwidth Statistics 200 500 1000 Mean 18.928 20.522 22.518 Standard Dev 35.247 39.247 40.186 Minimum 0 0 0 1st percentile 0.19 0.248 0.289 5th percentile 1.065 1.27 1.49 10th percentile 2.187 2.553 2.987 25th percentile 5.644 6.455 7.574 50th percentile 12.187 13.516 15.853 75th percentile 21.956 23.618 26.695 90th percentile 37.9 39.872 42.114 95th percentile 55.831 59.05 61.046 99th percentile 122.055 133.425 139.354 Maximum 4145.135 5116.346 5575.472 Table 7b: Summary Statistics for Percent Absolute Prediction Error: MA Data, TGWR Bandwidth Statistics 200 500 Mean 18.291 20.824 Standard Dev 32.995 35.477 Minimum 0.000 0.000 1st percentile 0.175 0.268 5th percentile 0.989 1.34 10th percentile 2.065 2.716 25th percentile 5.386 6.883 50th percentile 11.566 14.235 75th percentile 20.315 23.704 90th percentile 32.729 35.617 95th percentile 49.104 51.699 99th percentile 155.713 166.152 Maximum 1384.39 1533.289 Note for Table 7b: Some estimates here differ slightly from those in Table 5 because sample sizes differ due to being conditional on observations where PAPE is observed for all methods. In Table 7b it is TGWR for 200 and 500 bandwidths. In the MA results in Table 5, it is OLS, GWR and TGWR for 200 bandwidth. ===== PAGE 25 (score 0, numeric-paren 0, table-words 0, var-words 0) ===== 24 Figure 1: COD Values – OLS Figure 2: COD Values – GWR (200)