ML12340A815

Official Exhibit - NYS000446-00-BD01 - Analysis by Stephen C. Sheppard Using Tolley Mls Linear Square Root

ML12340A815
Person / Time
Site:	Indian Point
Issue date:	10/22/2012
From:	Sheppard S C - No Known Affiliation
To:	Atomic Safety and Licensing Board Panel
SECY RAS
References
RAS 23665, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML12340A815 (4)

Text

User: Stephen Sheppard Project: IPEC Statistics/Data Analysis name:

<unnamed>log:/home/steve/Documents/IndianPoint/Contention 17/TolleyOriginalLInearSquareRoot.smcl log type: smcl1. use "/home/steve/Documents/IndianPoint/Contention 17/MLS Data STATA Format.dta"

2.

• Tolley MLS model from report

3. regress askprice ipdist ipdistsq med_income house_age attached rail_dist pilotpay_2011 Source SS df MS Number of obs =

296 F( 7, 288) =

26.44 Model 7028465.17 7 1004066.45 Prob > F =

0.0000 Residual 10937065.5 288 37975.922 0.3912 0.3764 Total 17965530.7 295 60900.1041 Root MSE =

194.87 askprice Coef. Std. Err. t P>ltl [95% Conf. Interval]

ipdist ipdistsq med_income 2.382077 .68255 3.49 0.001 1.038658 3.725496 house_age attached rail_dist pilotpa~2011

_cons 577.5368 121.6523 4.75 0.000 338.0965 816.97714.

• Tolley model with IPEC impact modeled as linear in distance5. regress askprice ipdist med_income house_age attached rail_dist pilotpay_2011 Source SS df MS Number of obs =

296 F( 6, 289) =

30.03 Model 6899132.97 6 1149855.5 Prob > F =

0.0000 Residual 11066397.7 289 38292.0337 0.3840 0.3712 Total 17965530.7 295 60900.1041 Root MSE =

195.68 askprice Coef. Std. Err. t P>ltl [95% Conf. Interval]

ipdist 46.8946 14.10106 3.33 0.001 19.14079 74.6484 med_income 2.169724 .6755748 3.21 0.001 .8400531 3.499394 house_age attached rail_dist pilotpa~2011

_cons 409.3269 80.90215 5.06 0.000 250.0948 568.5596.

• Tolley model with IPEC impact modeled as proportional to square root of distance7. generate distsquareroot=ipdist^0.5

8. regress askprice distsquareroot med_income house_age attached rail_dist pilotpay_2011 Source SS df MS Number of obs =

296 F( 6, 289) =

29.64 Model 6843474.45 6 1140579.08 Prob > F =

0.0000 Residual 11122056.3 289 38484.6238 0.3809 0.3681 Total 17965530.7 295 60900.1041 Root MSE =

196.17 United States Nuclear Regulatory Commission Official Hearing Exhibit In the Matter of

Entergy Nuclear Operations, Inc. (Indian Point Nuclear Generating Units 2 and 3)

ASLBP #:07-858-03-LR-BD01 Docket #:05000247 l 05000286 Exhibit #:

Identified:

Admitted: Withdrawn:

Rejected: Stricken: Other: NYS000446-00-BD0110/22/2012 10/22/2012 Ori g inall y Identified as BRD000005Dr.To ll ey's original m odelModelw i thIPECimpact

proportional tolinear distance.NYS000446 Submitted: October 22, 2012 askprice Coef. Std. Err. t P>ltl [95% Conf. Interval]

distsquare~t 149.1657 48.24852 3.09 0.002 54.20262 244.1287 med_income 2.131092 .6773751 3.15 0.002 .7978782 3.464306 house_age attached rail_dist pilotpa~2011

_cons 296.0152 105.2117 2.81 0.005 88.9368 503.09369.

• Tolley model with IPEC impact modeled as proportional to square of distance10. regress askprice ipdistsq med_income house_age attached rail_dist pilotpay_2011 Source SS df MS Number of obs =

296 F( 6, 289) =

30.60 Model 6979449.73 6 1163241.62 Prob > F =

0.0000 Residual 10986081 289 38014.1211 0.3885 0.3758 Total 17965530.7 295 60900.1041 Root MSE =

194.97 askprice Coef. Std. Err. t P>ltl [95% Conf. Interval]

ipdistsq 7.8997 2.169942 3.64 0.000 3.628806 12.17059 med_income 2.254154 .6735366 3.35 0.001 .9284944 3.579813 house_age attached rail_dist pilotpa~2011

_cons 467.6234 73.78841 6.34 0.000 322.3925 612.8542Modelw i thIPECimpact

proportional tosquareroot ofdistance.Modelw i thIPECimpact

proportional tothesquare ofdistance.

User: Stephen Sheppard Project: IPEC Statistics/Data Analysis name:

<unnamed>log:/home/steve/Documents/IndianPoint/Contention 17/RepeatSalesAnalysisOfTolleyQuestions.s m log type: smcl1.

• Here is the model that is the basis of Dr. Sheppard's analysis:

2.Linear regression Number of obs =

1511 F( 2, 506) =

9.07 Prob > F =

0.0001 0.0076 Root MSE = .19957 (Std. Err. adjusted for 507 clusters in id) Robust nomreturn Coef. Std. Err. t P>ltl [95% Conf. Interval]

salepre74~76 distkm

_cons

.1585513 .0196089 8.09 0.000 .1200264 .19707623.

• Dr. Tolley raises a question about inclusion of data where one or more sale 4.

• involved a vacant lot. Consider the impact of excluding these observations

5.

• from the data used for the estimates:6. generate salewithlot=0

7. replace salewithlot=1 if lot==1 (325 real changes made)8.Linear regression Number of obs =

1222 F( 2, 414) =

11.33 Prob > F =

0.0000 0.0085 Root MSE = .18151 (Std. Err. adjusted for 415 clusters in id) Robust nomreturn Coef. Std. Err. t P>ltl [95% Conf. Interval]

salepre74~76 distkm

_cons

.1519886 .0211621 7.18 0.000 .1103901 .19358719.

• This shows that excluding these sales strengthens the results used in Dr. 10.

• Sheppard's analysis. The negative impact of the treatment group relative to the

11.

• control is significantly larger. The results are estimated with greater

12.

• precision.13.

• Dr. Tolley objects to including sales that occurred during one of the times of rapid 14.

• house price increase. Dropping these observations altogether is unwarranted.

15.

• If these time periods are different the preferred approach is to include an indicator Dr.S h eppar d's m odel.Modelw i thanyvacant lotdata excluded.

16.

• or dummy variable in the model to account for any excess returns. 17.

• We use indicator variables for the 1984Q2 to 1988Q1 time period and a separate

18.

• indicator for the time from 1999 through 2009.19> , vce(cluster id)

Linear regression Number of obs =

1222 F( 4, 414) =

11.19 Prob > F =

0.0000 0.0251 Root MSE = .18013 (Std. Err. adjusted for 415 clusters in id) Robust nomreturn Coef. Std. Err. t P>ltl [95% Conf. Interval]

salepre74~76 distkm dummy_80sb~e

.0653683 .023289 2.81 0.005 .0195888 .1111477 after98

.0538824 .0133926 4.02 0.000 .0275563 .0802085

_cons

.1123363 .0226813 4.95 0.000 .0677514 .156921220. *This shows that the result of Dr. Sheppard's analysis remains essentially 21.

• unaffected by accounting for the two time periods with rapid house price appreciation.

22.

• The time periods DO show unusually high returns to owning housing, but the impact of

23.

• the IPEC treatment is not statistically different from Dr. Sheppard's original result.

24.

• Modelw i thvacantlot dataexcluded andindicator variablesfor 1984-88and

1999-2009.