Do sunk costs affect prices in the housing market?
Dimuthu Ratnadiwakara and Vijay Yerramilli
30 Dec 2020
rm(list=ls())
library(stargazer)
library(plyr)
library(lfe)
library(zoo)
library(scales)
library(rdrobust)
library(data.table)
library(ggplot2)
library(dplyr)
library(fst)
library(fixest)zdata[,Selling_premium:=zdata$sales_price/zdata$adj_purch_price]
zdata[,Renter_fraction:=zdata$renters/zdata$totalpopulation]
zdata[,Purchase_price:=SalesPriceAmount_prev]
zdata[,Listing_price:=listing_amount]
zdata[,Selling_price:=sales_price]
zdata[,Listing_premium:=list_premium]
zdata[,Property_tax_last_year:=prop_tax_prev_year]
zdata[,Property_tax_rate_last_year:=prop_tax_rate]
zdata_reg <- zdata[PurchaseYear<2008 & ListingYear>2013 & house_age>0 ]
zdata_reg[,Time_on_market:=as.numeric(sale_date)-as.numeric(listed_date)+1]
zdata_reg[,zip_list_month:=paste(zip,ListingMonth)]
zdata_reg[,tract_list_month:=paste(GEOID,ListingMonth)]
setorder(zdata_reg,listed_date)zdata_reg[,predprice:= 0]
zd <- NULL
listmonths <- unique(zdata_reg$ListingMonth)
# pb <- txtProgressBar(min=1, max=length(listmonths), style=3)
for(i in 1:length(listmonths)) {
# setTxtProgressBar(pb, i)
temp <- CA_all[Sale_month < listmonths[i] & Sale_month> as.Date(paste0((year(listmonths[i])-1),"-",month(listmonths[i]),"-01")) & hpi_inflation>0 & sqft>0 & house_age>0 & !is.na(Purchase_price) & !is.na(hpi_inflation) & !is.na(beds) & !is.na(baths) & !is.na(sqft) & !is.na(house_age) & !is.na(yr) & !is.na(sales_price) & sales_price>0 & LotSizeSquareFeet>1000 & LotSizeSquareFeet<500000 & sales_price<2000000 & sales_price>100000 & adj_purch_price<2000000 & adj_purch_price>100000 & abs(sales_price-adj_purch_price) < 100000 ]
temp[,adjbin:=ntile(adj_purch_price,3)]
adjbinmaxmin <- temp[,.(minap=min(adj_purch_price),maxap=max(adj_purch_price)),by=adjbin]
t <- zdata_reg[ListingMonth==listmonths[i]]
t[,adjbin:=ifelse(adj_purch_price < adjbinmaxmin[adjbin==1]$maxap,1,
ifelse(adj_purch_price < adjbinmaxmin[adjbin==2]$maxap,2,3))]
# ifelse(adj_purch_price < adjbinmaxmin[adjbin==3]$maxap,3,
# ifelse(adj_purch_price < adjbinmaxmin[adjbin==4]$maxap,4,5))))]
if(nrow(temp)<30) {
t[,predprice:=adj_purch_price]
} else {
hedonicmodel <- feols(data=temp,sales_price~adj_purch_price+I(adj_purch_price^2)+I(adj_purch_price^3)+beds*factor(adjbin)+baths*factor(adjbin)+sqft*factor(adjbin)+log(house_age)*factor(adjbin)+log(LotSizeSquareFeet)*factor(adjbin))
t[,predprice:=(predict(hedonicmodel,newdata = t))]
}
zd <- rbind(zd,t)
}
zdata_reg <- zd
rm(zd)# zdata_reg[,predprice:=adj_purch_price]#exp(predict(hedonicmodel,newdata = zd))]
# zdata_reg[,predprice:=(predict(hedonicmodel,newdata = zdata_reg))]#
# zdata_reg[,predprice:=exp(predict(hedonicmodel,newdata = zdata_reg))]#
zdata_reg[,nominalloss:=Purchase_price-predprice]
zdata_reg[,nominalloss:=ifelse(nominalloss<=0,1,nominalloss)]
zdata_reg[,lppp:=listing_amount/predprice]
zdata_reg[,sppp:=sales_price/predprice]
zdata_reg <- zdata_reg[abs(predprice-sales_price)<400000]
zdata_reg[,lp:=floor(listing_amount/1000)*1000]
temp <- zdata_reg[,.(predprice=mean(predprice,na.rm=T),adj_purc=mean(adj_purch_price,na.rm=T),.N),by=lp]Figure 2
# ggplot(temp[N>10 & lp<500000],aes(x=lp,y=adj_purc))+geom_point()+geom_smooth(method="lm")+theme_minimal()+labs(x="Listing price",y="Adjusted purchase price")+ scale_x_continuous(labels = scales::comma)+ scale_y_continuous(labels = scales::comma)+labs(title="Adjusted purchase price")
ggplot(temp[N>10 & lp<1000000 & lp>=300000],aes(x=lp,y=predprice))+geom_point()+theme_minimal()+labs(x="Listing price ($)",y="Predicted price ($)")+ scale_x_continuous(labels = scales::comma)+ scale_y_continuous(labels = scales::comma)#+labs(title="Predicted price")
ggplot(temp[N>10 & lp<1000000 & lp>=300000],aes(x=lp,y=adj_purc))+geom_point()+theme_minimal()+labs(x="Listing price ($)",y="Adjusted purchase price ($)")+ scale_x_continuous(labels = scales::comma)+ scale_y_continuous(labels = scales::comma)#+labs(title="Predicted price")
Table 1
stargazer(zdata_reg[,c("Purchase_price","predprice","Listing_price","lppp","sppp","Property_tax_last_year","Property_tax_rate_last_year","purchase_hpi","hpi_inflation","ownership_years","beds","baths","sqft","LotSizeSquareFeet","house_age","LTV_prev","avg_school_rating","avg_school_distance","totalpopulation","medianage","medianhouseholdincome","Renter_fraction")],summary.stat = c("mean","sd","p25","median","p75","n"),digits = 4,
covariate.labels =c("Purchased price ($)","Predicted price ($)","Listing price ($)","Listing price/Predicted price","Selling price/Predicted price","Effective tax rate","Property taxes paid ($)","HPI_{Purchase}","HPI_{List}/HPI_{Purchase}","Years of ownership","Number of bedrooms","Number of bathrooms","House area (sq. ft)","Lot area (sq. ft)","Age of the house (years)","Loan-to-value_{Purchase}","GreatSchools rating","Distance to schools (miles)","Census tract population","Census tract median age","Census tract median income","Census traction fraction of renters") ,type="text" ) ###
## ===========================================================================================================
## Statistic Mean St. Dev. Pctl(25) Median Pctl(75) N
## -----------------------------------------------------------------------------------------------------------
## Purchased price () 447,057.3000 327,264.4000 225,000 363,000 575,000 24,869
## Predicted price () 589,070.6000 394,825.6000 322,428.2000 485,381.5000 723,442.5000 24,869
## Listing price () 648,591.2000 418,128.6000 385,000.0000 549,900.0000 779,000.0000 24,869
## Listing price/Predicted price 1.1625 0.3694 0.9560 1.0544 1.2179 24,869
## Selling price/Predicted price 1.1283 0.3546 0.9342 1.0250 1.1763 24,869
## Effective tax rate 5,608.9210 3,753.4250 3,114.6 4,713 7,010.6 24,869
## Property taxes paid () 0.0103 0.0037 0.0081 0.0099 0.0119 24,869
## HPIPurchase 420,132.7000 243,773.8000 246,500 381,100 541,900 24,869
## HPIList/HPIPurchase 1.4391 0.6173 0.9895 1.2447 1.7131 24,869
## Years of ownership 12.5796 2.7735 11 12 14 24,869
## Number of bedrooms 3.4142 0.8597 3 3 4 24,869
## Number of bathrooms 2.4665 0.9819 2.0000 2.0000 3.0000 24,869
## House area (sq. ft) 1,974.4900 810.8071 1,404 1,801 2,376 24,869
## Lot area (sq. ft) 16,926.5900 94,137.6500 5,652 7,200 10,010 24,869
## Age of the house (years) 39.1133 22.8273 19 35 56 24,869
## Loan-to-valuePurchase 0.6396 0.4074 0.3990 0.7845 0.8000 24,845
## GreatSchools rating 6.8483 2.0247 5.3333 7.0000 8.6667 24,843
## Distance to schools (miles) 1.4137 1.1366 0.8000 1.1333 1.6667 24,850
## Census tract population 5,880.3730 2,932.9850 4,123.5000 5,375.0000 6,801.0000 24,864
## Census tract median age 40.0679 7.8160 34.6000 39.2000 44.4000 24,862
## Census tract median income 80,951.6700 30,901.0600 57,917.0000 76,071.0000 99,583.0000 24,861
## Census traction fraction of renters 0.3230 0.1706 0.1924 0.2938 0.4301 24,862
## -----------------------------------------------------------------------------------------------------------Table 2
# stargazer(hedonicmodelfelm,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,covariate.labels = c("log(Purchase price)","log(HPI inflation)","Number of bedrooms","Number of bathrooms","House area","log(Age of the house)","log(Lot area)"), add.lines = list(c("Year", "Yes")),type="text")#Table 3
controls1 = "log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_list_month|"
controls2 = "log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|tract_list_month|"
controls3 = "log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction+log(nominalloss)|tract_list_month|"
controls4 = "log(adj_purch_price)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev|zip_list_month|"
cluster = "|zip"
instruments = "ownership_years" #
endo_var = "Property_tax_rate_last_year"regs <- list()
regs[[1]] <- felm(as.formula(paste("I(prop_tax_rate*100)~ownership_years+",controls1,"0",cluster,sep="")),data=zdata_reg)
regs[[2]] <- felm(as.formula(paste("I(prop_tax_rate*100)~ownership_years+",controls2,"0","|GEOID",sep="")),data=zdata_reg)
# .printtable(regs)
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = F,covariate.labels = c("Ownership years","log(Predicted price)","Number of bedrooms","Number of bathrooms","House area","GreatSchools rating","Distance to schools","Distance to amenities","log(Age of the house)","log(Lot area)","Loan-to-value_{Purchase}","Census tract median age","log(Census tract median income)","Census tract fraction of renters"),type="text")###
## ====================================================
## (1) (2)
## ----------------------------------------------------
## Ownership years -0.053*** -0.050***
## (0.001) (0.001)
## log(Predicted price) -0.702*** -0.761***
## (0.016) (0.021)
## Number of bedrooms 0.013*** 0.009
## (0.004) (0.007)
## Number of bathrooms 0.010*** 0.006
## (0.002) (0.008)
## House area 0.0002*** 0.0002***
## (0.00001) (0.00002)
## GreatSchools rating 0.013*** 0.013
## (0.003) (0.009)
## Distance to schools 0.003 -0.002
## (0.004) (0.011)
## Distance to amenities -0.0003* -0.001***
## (0.0001) (0.0003)
## log(Age of the house) -0.081*** -0.105***
## (0.008) (0.012)
## log(Lot area) 0.034*** 0.035***
## (0.006) (0.010)
## Loan-to-valuePurchase 0.038*** 0.036***
## (0.007) (0.011)
## Census tract median age 0.002**
## (0.001) (0.000)
## log(Census tract median income) 0.114***
## (0.020) (0.000)
## Census tract fraction of renters 0.078**
## (0.035) (0.000)
## N 23,811 23,811
## Adjusted R2 0.647 0.680
## ====================================================Table 5
olsformula <- as.formula(paste("log(Listing_price)~Property_tax_rate_last_year+",controls1,"0",cluster,sep=""))
ivformula <- as.formula(paste("log(Listing_price)~",controls1,"(",endo_var,"~",instruments,")",cluster,sep=""))
ivformula_tract <-as.formula(paste("log(Listing_price)~",controls2,"(",endo_var,"~",instruments,")","|GEOID",sep=""))
ivformula_tract_loss <-as.formula(paste("log(Listing_price)~",controls3,"(",endo_var,"~",instruments,")","|GEOID",sep=""))
regs <- list()
regs[[1]] <- felm(olsformula,data=zdata_reg)
regs[[2]] <- felm(ivformula,data=zdata_reg)
regs[[3]] <- felm(ivformula_tract,data=zdata_reg)
regs[[4]] <- felm(ivformula_tract_loss,data=zdata_reg)
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Effective tax rate","log(Predicted price)","Number of bedrooms","Number of bathrooms","House area","GreatSchools rating","Distance to schools","Distance to amenities","log(Age of the house)","log(Lot area)","Loan-to-value_{Purchase}","Census tract median age","log(Census tract median income)","Census tract fraction of renters","log(Nominal loss)"),
column.labels=c("OLS","IV"),column.separate=c(1,3),
add.lines = list(c("zipmonth", "Yes", "Yes", "No","No"),
c("tractmonth", "Yes", "Yes", "No","No"),
c("Cond. F. Stat","",round(condfstat(regs[[2]])[[1]],2),round(condfstat(regs[[3]])[[1]],2),round(condfstat(regs[[4]])[[1]],2))),type="text")###
## ============================================================================
## OLS IV
## (1) (2) (3) (4)
## ----------------------------------------------------------------------------
## Effective tax rate 15.934***
## (0.823)
## log(Predicted price) 0.350*** 0.307*** 0.299*** 0.293***
## (0.016) (0.016) (0.021) (0.023)
## Number of bedrooms 0.010*** 0.011*** 0.010** 0.010**
## (0.004) (0.004) (0.005) (0.005)
## Number of bathrooms 0.006 0.006 0.011* 0.011*
## (0.005) (0.005) (0.006) (0.006)
## House area 0.0002*** 0.0002*** 0.0002*** 0.0002***
## (0.00001) (0.00001) (0.00002) (0.00002)
## GreatSchools rating 0.013*** 0.014*** 0.012** 0.012**
## (0.003) (0.003) (0.006) (0.006)
## Distance to schools -0.004 -0.004 -0.003 -0.003
## (0.003) (0.003) (0.005) (0.005)
## Distance to amenities -0.0003*** -0.0003*** -0.001*** -0.001***
## (0.0001) (0.0001) (0.0002) (0.0002)
## log(Age of the house) -0.001 -0.008 -0.020** -0.020**
## (0.005) (0.006) (0.008) (0.008)
## log(Lot area) 0.059*** 0.060*** 0.055*** 0.055***
## (0.006) (0.006) (0.008) (0.008)
## Loan-to-valuePurchase 0.003 0.004 -0.001 -0.001
## (0.003) (0.003) (0.005) (0.005)
## Census tract median age 0.003*** 0.003***
## (0.0004) (0.0005) (0.000) (0.000)
## log(Census tract median income) 0.076*** 0.083***
## (0.012) (0.012) (0.000) (0.000)
## Census tract fraction of renters 0.054*** 0.058***
## (0.020) (0.021) (0.000) (0.000)
## log(Nominal loss) 0.0003
## (0.001)
## `Property_tax_rate_last_year(fit)` 10.667*** 11.441*** 10.836***
## (0.958) (1.449) (1.669)
## zipmonth Yes Yes No No
## tractmonth Yes Yes No No
## Cond. F. Stat 228.57 124.37 90.41
## N 23,811 23,811 23,811 23,811
## Adjusted R2 0.946 0.945 0.953 0.953
## ============================================================================Table 6
zhvi <- readRDS("C:/Users/dratnadiwakara2/Documents/OneDrive - Louisiana State University/Raw Data/Zillow Research Data/Zip_Zhvi_SingleFamilyResidence.rds")
zhvi <- data.table(zhvi)
zhvi <- zhvi[state=="CA"]
zhvi <- zhvi[month %in% as.Date(c("2006-01-01","2014-01-01"))]
zhvi <- zhvi[,c("zipcode","month","zhvi")]
zhvi <- dcast(zhvi,zipcode~month,value.var = "zhvi")
names(zhvi) <- c("zipcode","val2006","val2014")
zhvi <- zhvi[complete.cases(zhvi),]
zhvi <- data.table(zhvi)
zhvi[,pricechange:=log(val2014/val2006)]
hist <- ggplot()+geom_histogram(data=zhvi,aes(x=pricechange),fill="dodgerblue4",color="dodgerblue4",alpha=0.75)+labs(x="Price Change 2006 to 2014",y="Number of Zipcodes")+ theme_minimal()+scale_y_continuous(labels = comma)
zhvi[,pricechangeqt:=ntile(pricechange,16)]
zhvi <- zhvi[,c("zipcode","pricechange","pricechangeqt")]
zhvi[,zipcode:=as.numeric(zipcode)]
zdata_reg <- merge(zdata_reg, zhvi, by.x="zip",by.y="zipcode",all.x=T)
# print(hist)regs <- list()
regs[[1]] <- felm(olsformula,data=zdata_reg[nominalloss>1])
regs[[2]] <- felm(olsformula,data=zdata_reg[nominalloss<=1])
regs[[3]] <- felm(olsformula,data=zdata_reg[pricechangeqt<=4])
regs[[4]] <- felm(olsformula,data=zdata_reg[pricechangeqt %in% 5:16])
regs[[5]] <- felm(ivformula,data=zdata_reg[pricechangeqt<=4])
regs[[6]] <- felm(ivformula,data=zdata_reg[pricechangeqt %in% 5:16])
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Effective tax rate","log(Predicted price)"),
column.labels=c("OLS","IV"),column.separate=c(4,2),
add.lines = list(c("Controls", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("zipmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("Cond. F. Stat","","","","",round(condfstat(regs[[5]])[[1]],2),round(condfstat(regs[[6]])[[1]],2))),
keep=c("Property_tax_rate_last_year","predprice"),type="text")###
## =============================================================================================
## OLS IV
## (1) (2) (3) (4) (5) (6)
## ---------------------------------------------------------------------------------------------
## Effective tax rate 24.318*** 15.235*** 15.043*** 16.249***
## (2.189) (0.884) (1.954) (0.885)
## log(Predicted price) 0.639*** 0.314*** 0.451*** 0.338*** 0.460*** 0.291***
## (0.040) (0.016) (0.049) (0.017) (0.074) (0.016)
## `Property_tax_rate_last_year(fit)` 15.767*** 9.932***
## (4.012) (0.982)
## Controls Yes Yes Yes Yes Yes Yes
## zipmonth Yes Yes Yes Yes Yes Yes
## Cond. F. Stat 10.45 233.86
## N 5,451 18,360 3,776 20,015 3,776 20,015
## Adjusted R2 0.946 0.949 0.861 0.932 0.861 0.931
## =============================================================================================Table 7
regs <- list()
regs[[1]] <- felm(ivformula,data=zdata_reg[Listing_price<listing_q2])
regs[[2]] <- felm(ivformula,data=zdata_reg[Listing_price>listing_q2])
CA <- read_fst("C:/Users/dratnadiwakara2/Documents/OneDrive - Louisiana State University/Projects/Sunk Cost/Processed Data/all_homes_prev/CA_all_homes_prev.fst",columns = c("BuildingAreaSqFt","zip","Bedrooms"),as.data.table = T)
CA_summary <- CA[,.(
sqft_lb=quantile(BuildingAreaSqFt,0.25, na.rm=T),
sqft_hb=quantile(BuildingAreaSqFt,0.75, na.rm=T)),
by=zip]
temp <- CA[,.(
beds_lb=quantile(Bedrooms,0.25, na.rm=T),
beds_hb=floor(quantile(Bedrooms,0.75, na.rm=T)+0.99)),
by=zip]
CA_summary <- merge(CA_summary,temp,by="zip")
zdata_reg <- merge(zdata_reg,CA_summary,by="zip",all.x=T)
zdata_reg[,outlider_sqft:=ifelse(sqft>sqft_hb ,1,0)]
zdata_reg[,outlider_beds:=ifelse(beds>beds_hb,1,0)]
regs[[3]] <- felm(ivformula,data=zdata_reg[outlider_sqft==0 & outlider_beds==0])
regs[[4]] <- felm(ivformula,data=zdata_reg[outlider_sqft==1 | outlider_beds==1])
redfin <- read.csv("C:/Users/dratnadiwakara2/Documents/Research/UH computer27/sunkcost_2019/home_sales_redfin.csv")
redfin$month <- as.Date(redfin$month,origin="1900-01-01")
redfin <- data.table(redfin)
redfin <- redfin[redfin$month<="2016-01-01"]
redfin <- redfin[,.(median_home_sales=median(home_sales)),by=list(zip)]
redfin[,low_activiy:=ifelse(redfin$median_home_sales<quantile(redfin$median_home_sales,0.75),1,0)]
names(redfin) <- c("zip","median_home_sales_2","low_activity_2")
zdata_reg <- merge(zdata_reg,redfin,by="zip",all.x=TRUE)
regs[[5]] <- felm(ivformula,data=zdata_reg[low_activiy==1])
regs[[6]] <- felm(ivformula,data=zdata_reg[low_activiy==0])
temp <- zdata_reg[!duplicated(zdata_reg$GEOID)]
median_age <- median(temp$medianage,na.rm = TRUE)
regs[[7]] <- felm(ivformula,data=zdata_reg[medianage<=median_age])
regs[[8]] <- felm(ivformula,data=zdata_reg[medianage>median_age])
medincome <- median(temp$medianhouseholdincome,na.rm = TRUE)
regs[[9]] <- felm(ivformula,data=zdata_reg[medianhouseholdincome<=medincome])
regs[[10]] <- felm(ivformula,data=zdata_reg[medianhouseholdincome>medincome])
median_renterfrac <- median(temp$Renter_fraction,na.rm = TRUE)
regs[[11]] <- felm(ivformula,data=zdata_reg[Renter_fraction<=median_renterfrac])
regs[[12]] <- felm(ivformula,data=zdata_reg[Renter_fraction>median_renterfrac])
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Effective tax rate"),
column.labels=c("L Pr.","H Pr","L Sz","H Sz","L Act","H Act","L Ag","H Ag","L Inc","H Inc","L Ren","H Ren"),
add.lines = list(
c("Controls", rep("Yes",12)),
c("zipmonth", rep("Yes",12)),
c("Cond. F. Stat",round(c(condfstat(regs[[1]])[[1]],condfstat(regs[[2]])[[1]],condfstat(regs[[3]])[[1]],condfstat(regs[[4]])[[1]],condfstat(regs[[5]])[[1]],condfstat(regs[[6]])[[1]],condfstat(regs[[7]])[[1]],condfstat(regs[[8]])[[1]],condfstat(regs[[9]])[[1]],condfstat(regs[[10]])[[1]],condfstat(regs[[11]])[[1]],condfstat(regs[[12]])[[1]]),2))),
keep=c("Property_tax_rate_last_year"), flip=T,
type="text")###
## =====================================================================================================================================
## L Pr. H Pr L Sz H Sz L Act H Act L Ag H Ag L Inc H Inc L Ren H Ren
## (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
## -------------------------------------------------------------------------------------------------------------------------------------
## Effective tax rate 4.309*** 13.252*** 8.301*** 16.460*** 10.972*** 9.702*** 6.127*** 12.732*** 15.873*** 8.664*** 10.036*** 12.662***
## (1.068) (1.466) (1.093) (2.073) (1.457) (1.260) (1.498) (1.310) (1.999) (1.054) (1.282) (1.589)
## Controls Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
## zipmonth Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes
## Cond. F. Stat 119.18 96.6 160.47 79.5 114.17 134.81 58.39 162.66 41.79 215.48 146.25 71.22
## N 12,257 11,440 15,487 8,324 10,275 12,269 9,297 14,514 8,419 15,392 15,312 8,499
## Adjusted R2 0.959 0.956 0.944 0.947 0.937 0.943 0.941 0.941 0.922 0.939 0.948 0.939
## =====================================================================================================================================Table 8 (Panel A)
regs <- list()
regs[[1]] <- felm(as.formula(paste("log(sales_price)~Property_tax_rate_last_year+",controls1,"0",cluster,sep="")),data=zdata_reg)
regs[[2]] <- felm(as.formula(paste("log(sales_price)~",controls1,"(",endo_var,"~",instruments,")",cluster,sep="")),data=zdata_reg)
regs[[3]] <- felm(as.formula(paste("log(Time_on_market)~Property_tax_rate_last_year+",controls1,"0",cluster,sep="")),data=zdata_reg)
regs[[4]] <- felm(as.formula(paste("log(Time_on_market)~",controls1,"(",endo_var,"~",instruments,")",cluster,sep="")),data=zdata_reg)
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Effective tax rate","log(Predicted price)"),
column.labels=c("log(Sales price)","log(Days on market)"),column.separate=c(2,2),
add.lines = list(c("Controls", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("zipmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("Cond. F. Stat","",round(condfstat(regs[[2]])[[1]],2),"",round(condfstat(regs[[4]])[[1]],2))),
keep=c("Property_tax_rate_last_year","predprice"), type="text")# ##
## =========================================================================
## log(Sales price) log(Days on market)
## (1) (2) (3) (4)
## -------------------------------------------------------------------------
## Effective tax rate 16.089*** -0.449
## (0.845) (1.778)
## log(Predicted price) 0.349*** 0.306*** -0.039* 0.022
## (0.016) (0.017) (0.023) (0.032)
## `Property_tax_rate_last_year(fit)` 10.789*** 7.069**
## (1.038) (3.166)
## Controls Yes Yes Yes Yes
## zipmonth Yes Yes Yes Yes
## Cond. F. Stat 228.57 228.57
## N 23,811 23,811 23,811 23,811
## Adjusted R2 0.920 0.919 0.342 0.341
## =========================================================================temp <- zdata_reg
temp[,purch_month_no:=as.numeric(format(temp$PurchaseMonth,"%m"))]
temp[,sale_month_no:=as.numeric(format(temp$sale_date,"%m"))]
temp <- temp[!is.na(temp$predprice) & !is.na(temp$GEOID)]
temp[,zip_purch_list:=paste(temp$zip,temp$purchase_list_year)]
adj_proptax <- felm(Property_tax_rate_last_year~log(predprice)|zip_purch_list,data=temp)
adj_proptax_resid <- data.frame(adj_proptax$residuals)
names(adj_proptax_resid) <- c("adj_proptax_resid")
adj_list <- felm(log(Listing_price)~log(predprice)|zip_purch_list,data=temp)
adj_list_resid <- data.frame(adj_list$residuals)
names(adj_list_resid) <- c("adj_list_resid")
adj_sale <- felm(log(Selling_price)~log(predprice)|zip_purch_list,data=temp)
adj_sale_resid <- data.frame(adj_sale$residuals)
names(adj_sale_resid) <- c("adj_sale_resid")
adj_purch <- felm(log(Purchase_price)~0|zip_purch_list,data=temp)
adj_purch_resid <- data.frame(adj_purch$residuals)
names(adj_purch_resid) <- c("adj_purch_resid")
temp <- cbind(temp,adj_proptax_resid)
temp <- cbind(temp,adj_list_resid)
temp <- cbind(temp,adj_sale_resid)
temp <- cbind(temp,adj_purch_resid)
temp[,DocumentDate_prev:=as.Date(temp$DocumentDate_prev,origin = "1970-01-01")]
temp[,purch_week:=as.numeric(strftime(temp$DocumentDate_prev,format="%V"))-20]
temp[,postJune:=ifelse(purch_week>-2,1,0)]
temp[,zip_purch_year:=paste(zip,PurchaseYear)]Table 9
dscregs <- list()
dscregs[[1]] <- felm(log(Purchase_price)~postJune+purch_week|zip_purch_year|0|zip,data=temp)
dscregs[[2]] <- felm(beds~postJune+purch_week|zip_purch_year|0|zip,data=temp)
dscregs[[3]] <- felm(baths~postJune+purch_week|zip_purch_year|0|zip,data=temp)
dscregs[[4]] <- felm(log(sqft)~postJune+purch_week|zip_purch_year|0|zip,data=temp)
dscregs[[5]] <- felm(log(house_age)~postJune+purch_week|zip_purch_year|0|zip,data=temp)
# dscregs[[6]] <- felm(log(0.0001+LTV_prev)~postJune|zip_purch_year|0|zip,data=temp)
stargazer(dscregs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = F,
covariate.labels = c("Purchased after June 01"),
add.lines = list(c("zippurchmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes")),
column.labels=c("log(Purchase price)","No beds","No baths","log(House area)","log(Age of house)","log(Loan-to-value)"),type = "text")##
## ===============================================================================================
## log(Purchase price) No beds No baths log(House area) log(Age of house)
## (1) (2) (3) (4) (5)
## -----------------------------------------------------------------------------------------------
## Purchased after June 01 0.035*** -0.043** -0.037* -0.027*** 0.025**
## (0.010) (0.020) (0.020) (0.010) (0.013)
## purch_week 0.0005 0.001 0.0001 0.0003 -0.001
## (0.0003) (0.001) (0.001) (0.0004) (0.0004)
## zippurchmonth Yes Yes Yes Yes Yes
## N 24,864 24,864 24,864 24,864 24,864
## Adjusted R2 0.651 0.174 0.143 0.190 0.485
## =============================================================================================== # keep=c("Property_tax_rate_last_year","predprice"),
# type="text")#Table 10 (Panel A)
regs = list()
regs[[1]] = felm(Property_tax_rate_last_year~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
regs[[2]] = felm(log(Listing_price)~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
regs[[3]] = felm(log(Selling_price)~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Purchased after June 01","Week of purchase","log(Predicted price)"),
column.labels=c("Effective tax rate","log(Listing price)","log(Selling price)"),
add.lines = list(c("Controls", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("zippurchmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes")),
keep=c("postJune","predprice","purch_week"),type="text")###
## ================================================================================
## Effective tax rate log(Listing price) log(Selling price)
## (1) (2) (3)
## --------------------------------------------------------------------------------
## Purchased after June 01 -0.0002*** -0.012*** -0.011**
## (0.0001) (0.004) (0.005)
## Week of purchase 0.00001*** 0.0004*** 0.0004**
## (0.00000) (0.0001) (0.0002)
## log(Predicted price) -0.007*** 0.240*** 0.246***
## (0.0002) (0.014) (0.014)
## Controls Yes Yes Yes
## zippurchmonth Yes Yes Yes
## N 23,811 23,811 23,811
## Adjusted R2 0.670 0.941 0.934
## ================================================================================Table 10 (Panel B)
regs = list()
regs[[1]] = felm(Property_tax_rate_last_year~postJune+purch_week+I(purch_week^2)+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
regs[[2]] = felm(log(Listing_price)~postJune+purch_week+I(purch_week^2)+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
regs[[3]] = felm(log(Selling_price)~postJune+purch_week+I(purch_week^2)+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp)
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Purchased after June 01","Week of purchase","Week of purchase^{2}","log(Predicted price)"),
column.labels=c("Effective tax rate","log(Listing price)","log(Selling price)"),
add.lines = list(c("Controls", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("zippurchmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes")),
keep=c("postJune","predprice","purch_week"),type="text")#, type="text"##
## ================================================================================
## Effective tax rate log(Listing price) log(Selling price)
## (1) (2) (3)
## --------------------------------------------------------------------------------
## Purchased after June 01 -0.0002** -0.011** -0.010*
## (0.0001) (0.005) (0.006)
## Week of purchase 0.00001*** 0.0004 0.0003
## (0.00000) (0.0002) (0.0002)
## Week of purchase2 -0.00000 0.00000 0.00000
## (0.00000) (0.00001) (0.00001)
## log(Predicted price) -0.007*** 0.240*** 0.246***
## (0.0002) (0.014) (0.014)
## Controls Yes Yes Yes
## zippurchmonth Yes Yes Yes
## N 23,811 23,811 23,811
## Adjusted R2 0.670 0.941 0.934
## ================================================================================Table 10 (Panel C)
Local linear running variable (3-months)
regs = list()
regs[[1]] = felm(Property_tax_rate_last_year~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp[abs(purch_week)<=12])
regs[[2]] = felm(log(Listing_price)~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp[abs(purch_week)<=12])
regs[[3]] = felm(log(Selling_price)~postJune+purch_week+log(predprice)+beds+baths+sqft+avg_school_rating+avg_school_distance+walk_score+log(house_age)+log(LotSizeSquareFeet)+LTV_prev+medianage+log(medianhouseholdincome)+Renter_fraction|zip_purch_list|0|zip,data=temp[abs(purch_week)<=12])
stargazer(regs,no.space = T,align = T,omit.stat=c("ser","f", "rsq"),style = "qje",omit.table.layout = "n", dep.var.labels.include = FALSE,
covariate.labels = c("Purchased after June 01","Week of purchase","log(Predicted price)"),
column.labels=c("Effective tax rate","log(Listing price)","log(Selling price)"),
add.lines = list(c("Controls", "Yes", "Yes", "Yes","Yes","Yes","Yes"),c("zippurchmonth", "Yes", "Yes", "Yes","Yes","Yes","Yes")),
keep=c("postJune","predprice","purch_week"),type="text")#,type="text"##
## ================================================================================
## Effective tax rate log(Listing price) log(Selling price)
## (1) (2) (3)
## --------------------------------------------------------------------------------
## Purchased after June 01 -0.0002 -0.012* -0.015**
## (0.0001) (0.007) (0.007)
## Week of purchase 0.00001* 0.0004 0.001
## (0.00001) (0.001) (0.001)
## log(Predicted price) -0.007*** 0.233*** 0.238***
## (0.0002) (0.014) (0.016)
## Controls Yes Yes Yes
## zippurchmonth Yes Yes Yes
## N 12,580 12,580 12,580
## Adjusted R2 0.673 0.940 0.937
## ================================================================================